Eurokarst is the European bi-annual conference on the Hydrogeology of Karst and Carbonate Reservoirs.
Due to the COVID pandemics, the organisation of the next Eurokarst had to be postponed. It will be held in Malaga, June 22-25, 2022.
Eurokarst offers a platform for professional exchange between field practitioners and academic researchers. Eurokarst is organized usually every two years by the Universities of Neuchâtel (Switzerland), Besançon (France), and Málaga (Spain).
A multidisciplinary investigation was carried out in a karstic depression in a housing development under construction in Madrid to assess its stability. It was found that it is a small basin within a larger depression as a result of subsidence accumulated during the Quaternary. Subsidence has built up progressively in the Miocene clay cap and bedrock due to the underlying dissolution of gypsum rich intercalations. The preferential circulation of the dissolving subsurface flow is along a fault conditioned by subsidence, the formation of an elongated syncline along the fracture, and the alluvial basin. During the Pleistocene, a former lagoon zone was formed in this subsiding area; it was also a groundwater discharge zone. The decrease in recent times is very small and could be evaluated to be about 0.4 mm/year, and affects the alluvial zone and along the furrow of a fault zone, where the maximum average rate of subsidence would be 1.4 mm/year. This has led to the development of a relatively strong alluvium. It seems that under the alluvial deposits, a slow and diffuse dissolution is taking place of the shallower clayey gypsiferous levels, free of hypersoluble mineral species; this is somewhat more intense in the fault zone, which is more active hydrodynamically, where groundwater velocity is higher. Microgravimetry surveys indicate that only 5% of the area hidden under the alluvium shows anomalies, interpreted as residual soft clayey masses, or anomalous alluvial fillings of old dissolution voids. These pockets (“bolsones”), have dimensions of no more than 20 _ 20 m and depths below 20 m. These measurements have been confirmed by boreholes and are the only points that would require special attention in the future construction of the urbanization.
The urbanization work, in full development, is implementing solutions aimed at the stability of the road in the strips of alluvial studied.
The affected areas by this casuistry are in areas classified as green zones in the planning (parks and open spaces). However, and despite obtaining non-relevant subsidence data, exhaustive research works have been carried out, based on extensive field prospecting (trial pits, boreholes, micro gravimetry, etc.) These works, once analysed, have led to satisfactory conclusions for the proper development of the urbanization works in progress.
A new and simple proposal of karst architectural types and their geomorphological and hydrogeological implications: wardrobe-type karst and bed-type karst
Juan José Durán-Valsero, Eulogio Pardo-Igúzquiza, Sergio Raúl Durán-Laforet, Raquel Morales-García and Pedro Agustín Robledo-Ardila
Abstract
Ever since the first works of researchers at the beginning of the 20th century (Cvijic, Grund), there have been different classifications for all the types of karst from different regions of the world. These classifications were generally linked to dynamic perspectives of the geological time, establishing several evolutive stages of karst, according to the ideas linked to the evolution of the karstic relief according to “geomorphologic cycles” (young, adolescent, mature and old karst). In recent times, some authors (e.g., Veress, 2020) have proposed detailed classifications, with a hierarchy of groups, subgroups, types, and subtypes, defining up to 81 types of karst distributed throughout all regions of the planet. However, none of these classifications refer to the three-dimensional configuration of the karstic massifs, in other words, to their spatial architecture. Our very simple proposal consists of establishing two opposite conceptual models, based on the geometry of the karst, between which there may be other intermediate terms. These models can serve as a theoretical conceptual framework to explain the geomorphological characteristics of exokarst and endokarst (horizontal and vertical caves), as well as the karstic hydrogeological functioning. The two proposed conceptual models are the “wardrobe-type” karst and the “bed-type” karst. The first one corresponds to karsts with an important development in the vertical dimension (generally mountain or high-relief karsts), with significant altimetric differences between the recharge zones (that have a high concentration of typical exokarstic morphologies, such as dolines, poljes or lapiace fields) and the discharge zones (springs, linked to river valleys, lakes, or the sea). In this type of karst, the unsaturated zone has a very important development, reaching hundreds or even thousands of meters below the topographic surface, conditioning the presence of large vertical cavities and the hydrodynamics of the hydrological system. On the contrary, in bed-type karsts, the vertical dimension is much more attenuated, being the horizontal dimension more important, reaching a greater superficial extension of the karsts. The water table is usually relatively close to the surface (meters to tens of meters, generally) and, consequently, the unsaturated zone presents little to no development. The exokarstic morphologies are distributed throughout the surface of the karst and the underground systems present a pattern where horizontal developments are more important (caves, multilevel systems, etc.). Most of the world’s great karsts can be associated with one of these two extreme models: the mountains of the Western Caucasus (where the deepest vertical caves of the planet are found) or the Picos de Europa (Spain) are good examples of the wardrobe-type karsts, whereas the Yucatán Peninsula (Mexico-Guatemala-Belize) or Campo de Montiel-Lagunas de Ruidera (Spain) are examples of bed-type karsts.
Los estudios de petrografía de los espeleotemas hacen posible determinar los episodios de ruptura, interrupción del crecimiento o fábrica sedimentaria que se producen tanto durante su formación, como en momentos posteriores. Estas variaciones texturales observadas responden a las diferentes condiciones fisicoquímicas existentes en el medio durante su formación actuando como indicadores paleoclimáticos. El análisis petrográfico de un espeleotema de la Cueva de Nerja (Málaga, Sur de España) ha permitido identificar su fábrica cristalina, es decir, la morfología cristalina y la disposición espacial de los cristales primarios que lo forman, que está relacionada con la variabilidad del flujo de agua disponible, la química, la tasa de goteo o la tasa de desgasificación de CO2. Así mismo, se han reconocido las inclusiones fluidas que se forman durante el crecimiento de los cristales, constatándose en algunas de ellas restos del agua kárstica existente en la cueva durante su formación. El análisis de la relación genética entre la inclusión fluida y la calcita colindante permite definir la tipología de las inclusiones fluidas en función de la relación temporal con la calcita (primarias o secundarias), de la relación espacial con los cristales de calcita (inter- o intracristalinas) y de su morfología.
Ines EZZINEª, Amal MHIMDIª, Fadoua HAMZAOUIᵇ, Mohamed GHANMIª, Rachida BOUHLILAᶜ
ªLaboratory of Geoscience, Resource mineral, Energy and Environmental (LGREE), Geology Department, Faculty of Sciences of Tunis, University of Tunis El Manar, Tunis 1060, Tunisia.
E-mail : ezzineines82@gmail.com
ᵇLaboratory of Sedimentary Basins and Petroleum Geology (SBPG), LR18 ES07, Geology Department, Faculty of Sciences of Tunis, University of Tunis El Manar, Tunis 1060, Tunisia.
E-mail : fadoua_fst@yahoo.fr
ᶜDepartment of Civil Engineering, Modelling in Hydraulic and Environment, Laboratory, National Engineers School of Tunis, University of Tunis El Manar, Tunis, Tunisia.
E-mail : rachida.bouhlila@enit.utm.tn
Abstract:
The area located in Northern Africa, numerous existing caves in the Jebel Zaghouan (NE Tunisia) allow the practice of caving. They are the result of the very structure of this limestone massif fractured by numerous faults that allowed the water runoff to infiltrate the center of the massif. Located in South-West of Jebel Zaghouan, the Jebel Stah cave is called “ante-prehistoric” because its formation certainly predates the arrival of the oldest prehistoric hominids. Its access is relatively easy, it has the undeniable advantages of being a real cave and of allowing an initiation, without great difficulty to speleology. About 150 to 200 meters long, not very deep, but it offers visitors beautiful rooms decorated with magnificent concretions, on which the light of the lamps, reflected by the thin film of water that covers everything, lights up flashes and colors: from sparkling white, to beige, ochre, to rusty color, and creates fantastic, moving, almost living shadows.
We had the chance to visit the cave of Jebel Stah and to discover it. In fact, after having climbed the 900 meters of altitude in the mountain Jebel Stah in Zaghouan to arrive at the small hole which has leaded us to the cave. We did 10 meters of descent then a second one of 6 meters to reach the plateau of cave, we notice that there is no water in the cave, which indicate a water shortage in Jebel Zaghouan.
Keywords: Zaghouan, cave, Jebel Stah, speleology.
Karstic caves have historically been used as a tourist resource, due to their important cultural and natural values. In addition to its conventional tourist use, the underground environment has also been used as a place for the celebration of a wide range of activities, among which the celebration of concerts with relatively large public attendance of people stands out.
The development of these events can induced an additional environmental impact to those produced by the tourist visit. On the one hand, they require the installation of infrastructures such as the stage, sound systems, artistic lighting and seating for the public, that can damage the walls and speleothems (incisions, scratching...) when they are used as direct support. Moreover, the presence of these allochthonous objects in the cave favours the development of microorganisms, especially fungi, considered one of the main factor of rock art biodeterioration. On the other hand, events usually take place outside of tourist opening hours therefore, they imply a reduction in the cavity recovery time, necessary to remove the daily impact of the visits.
The fragile environmental balance of caves makes necessary to study and control this type of activities, as well as to design and implement specific conservation protocols aimed to prevent or minimize their impact. The Nerja Cave (Malaga, Spain) is an important touristic cave declared Good of Cultural Interest with the category of Archaeological Zone and an internationally relevant Site of Geological Interest (Global Geosite), where, an annual music and dance festival took place from 1959 to 2019. The enviromental monitoring of this festival along several years has allowed to know and value its impact from the cave conservation point of view and to tests the efficacy of the conservation protocols applied. The results obtained highlight the negative impact that these types of events can produce in the caves and, therefore, the convenience of developing specific studies to evaluate whether or not they should be celebrated inside them.
Canale di Pirro is the main polje in Apulia, extending for more than 12 km in W-E direction in the Low Murge karst area of the central sector of the region. Besides being a remarkable example of surface karst landscape, it hosts the deepest cave in the region (Inghiottitoio della Masseria Rotolo), one of the twos where cavers directly reach the water table at depth of -264 m below the ground surface. Speleodiving explorations added further 60 m, so that the total depth of the system is -324 m.
Geophysical analyses have been performed in the Canale di Pirro polje, in correspondence of two sites of particular interest. The first is a small sinkhole within a vineyard, with depth of about 3m, circular shape and diameter of some 7m. The survey was addressed to detect likely ways of water infiltration at the site, and the possible presence of underground voids and conduits in the limestone bedrock.
The second site (Gravaglione) is located about 0,5km upstream of the cave before mentioned, and is the main swallow hole in the polje. It is an elongated opening in the ground, where 4m of terra rossa deposits are exposed. During rainstorms, most of the water accumulated in this sector of the polje is drained at this site, with a time of absorption ranging from a few hours to some days, also in function of the saturation condition of the sediments. The survey was addressed to verify the existence of caves large enough to be explored by man, and therefore the possibility to open a passage in the cover deposits.
By adopting a multi-array and multi-scale strategy, based on the acquisition and joint inversion of geoelectrical data collected by using different arrays and with electrode spacing ranging from 5m to 20m, a high-resolution model of the electrical subsurface structure was obtained. With an investigation depth up to ~150 m below the ground, and a high surficial resolution, the model shows features compatible with the presence of large caves and provides useful information on their possible extension and on the thickness of the cover deposits.
Show caves are great natural attractions and constitute an important economic engine for a defined area. However, cave management requires knowledge of the cave dynamics to ensure an adequate exploitation and conservation (e.g. number of visitors, amount of CO2…). Show caves located close to the hydrological base level are usually affected by a sudden rise of the water level in response to rainfall events, exposing tourism activity to flood risk. Here we present a monitoring campaing in Las Güixas show cave, a small tourist cave located in Central Pyrenees, to help with the preparation of flood risk management. Around 28000 people year-1, organized in groups of 20-35 people during 1-hour, visit the cave. The cave is formed by three levels; the lower level keeps flooded by waters, and drains to the Aragon River; the middle level is connected with the lower level by a 15 m ramp and is open to tourism. Finally, the upper level is composed by several fossil galleries non-affected by hydrological activity during regular rainfalls. Cave monitoring includes temperature, relative humidity, CO2 and water level measurements to characterize the different cave sectors, as well as to assess the possible influence of the visitors on the cave dynamics. The touristic sector is very well ventilated due to the intense air flow associated to chimney effect, and therefore CO2 remains low along the year (around 410 ppm). Temperature shows high thermal oscillations, ranging between 3°C (February) and 17°C (August). During the maximum number of visits per day, the CO2 generated by the tourists increases to 550 ppm, however, is quickly reduced by the ventilation. Water level monitoring shows clear rises (up to 20 cm min-1) depending of the amount of rainfall and snow melting. The water level rise provokes the flood of the touristic sector where the discharge reaches 900 l s-1 after 30-60 mm rainfall events. However, the water level does not always respond to rainfall amount in the same way, indicating that the water retained in the lower level plays an important role. Once the touristic cave is flooded, the visits for the following days have to be cancelled. The cave remains closed to the public, even when the water level descents to the siphon area and the alert system remains active, as long as the rainfall forecast is bad. This situation may last for several weeks, which generates important economic losses. In addition, the higher water flow increases the natural inputs of CO2 due to degassing, which reaches 900 ppm in the cave atmosphere. Güixas cave monitoring shows that anthropic CO2 emissions remain substantially lower than during the floods. In the future, hydrological cave modelling will help to anticipate flood hazards and risk management actions during the rainfall season.
The present contribution compares the geomorphological characteristics of dolines developed in three different geological contexts in southern Spain: an evaporite tectonic melánge, a bare-carbonate massif, and an outcrop of interstratified sedimentary gypsum sequence. In the three cases, the closed depressions have been identified and mapped using digital elevation models with a spatial resolution of 5 m, and the results were validated in the field. To reveal similitudes and differences due to distinctive geological and climatological settings between the three sites, morphometric and spatial analyses were applied to the threes sites: size-distributions, relationships between area and depth, preferential directions, and point field analysis, among others. The results show apparent differences between the carbonate and the two evaporitic settings. Carbonate outcrop displays larger dolines, im-plying a lower doline density, while the density of depressions rises in the evaporite sites. The results also show a dominant ENE-WSW alignment of dolines in two areas, in agreement with the main fault families of the Betic Cordillera, which evidences the tectonic influence on the karst development. However, in the gypsum karst, other directions appear related to the surficial drainage network connected to dolines. Finally, in the tectonic melange, the uneven distribution of lithologies affects karstification distribution. Geological differences (both lithology and structure) imply changes in solubility and karstification dynamic that, together with other factors such as the climatic conditions and the exposure time, affect the genesis of the depression and, ultimately, its size and shape. All that combined explain the differences in morphometric parameters and spatial.
One-third of the world’s soils are in the tropics and approximately two-thirds of the world’s population lives in this region, but studies about soil thickness in the tropics are still scarce. Soil plays an important role in protecting aquifers and most of their vulnerability in the karst environment is due to soil thickness. Physical, chemical and biological soil processes involve many reactions that naturally attenuate the potential effect of contaminants, and these reactions are influenced by soil depth. This study sought to conduct in-depth characterization of soil thickness and its mineralogy, pH, texture, porosity and density. The study area is in the municipality of Sete Lagoas, Minas Gerais in southeastern-Brazil. Sixteen soil samples were taken from a representative profile, and a saprolite sample was taken from a point close to the profile. The soil is a typical highly weathered red Oxisol. The samples for mineralogy analyses were removed with a hand auger. The samples for porosity and density analyses were removed with a 50 x 50 mm steel ring sampler. The macro and microporosities were measured in a sandbox at a voltage of 6 kPa. The granulometry was measured by the pipette method. The magnetic fraction was sorted from the sand fraction with a hand magnet. The samples for XRD analysis were oven dried at 600 °C for 24 hours and ground with a porcelain pestle. They were analyzed with a Phillips-PANalytical diffractometer, using CuKα and monochromator radiation according to interplanar distances and peak intensities in the diffractograms, based on a reference sample, using the PDF-2 Release 2010 of the ICDD – International Center for Diffraction Date and X’Pert HighScore software. The XRF samples were pulverized and analyzed with an EDX-720 Shimadzu. The pH was measured with a pHmeter in a 1:2.5 soil/water proportion. The results showed remarkable mineralogical uniformity in the profile, to a depth of 8 m, with a small increase in the kaolinite content below 9 m down to 15 m, and a steep decrease in the kaolinite content in the saprolite. The same trends can be seen in the clay content and in the magnetic fraction. The geochemical indexes indicate, in addition to soil uniformity, a high level of soil weathering and its contrast with the saprolite. The soil density and the macro and microporosities also show a uniformity in the profile. Lowers levels of soil density and microporosity and higher levels of macroporosity close to the surface may be due to biological activity, as this area is still covered with native savanna. The chemical and physical results indicate that the soil profile is indeed very deep and uniform, and that the pedogenetic processes are active many meters below the surface. The levels of macroporosity throughout the profile indicate that this soil has no restriction to water percolation and may be an important aquifer recharge area.
Keywords: Soil, Thickness, Karst, Mineralogy, Brazil
Rull Cave is a karstic cavity located in Vall d’Ebo (Alicante), in the Southeast of Spain. The cave is developed in massive Miocene conglomerates and Cretaceous limestones. The cave presents a single rounded hall of 1535 m length with a 3 m2 entrance located in its highest part. It is a tourist cave with an average of 15550 visitors per year.
A detailed microenvironmental monitoring system has been measuring the environmental conditions in the cave indoors and outdoors, as well as the physical properties above the cave, from 2012 to 2021 to understand the gaseous dynamics and their relationships in this three-component system (soil-cave-atmosphere). Besides the presence of visitors, Rull Cave presents, every year, stable values of temperature (average value of 16.2 ºC, with variations of ±1.0ᵒC) and relative humidity (97.5%).
The microclimatic measurements of air cave reveal two different gaseous stages in an annual cycle. In the study period, maximum average values of CO₂ and ²²²Rn are reached in the hottest months (4000 ppm of CO₂ and 5600 Bq/m3, respectively), while coldest months are related to the lowest gaseous concentrations (480 ppm of CO₂ and 427 Bq/m3, respectively). The annual cycles described by CO₂ and ²²²Rn are consequence of the gaseous exchange between the outside atmosphere and the indoors cavity through the porous system and fissures of the soil and the host rock, controlled by the variations of the indoors and outdoors environmental variables.
Average value of atmospheric temperature and relative humidity are, respectively, 16.1 ºC, nearly coincident with indoors temperature, and 69.9%, with an average measured annual precipitation of 437 mm. Soil dynamics is strongly related to cave dynamics. Thus, soil conditions are also measured. For the study period, average soil temperature and volumetric water content are 16.6 ºC and 0.2.
Complementary discrete sampling of δ13C of CO₂ in cave, outdoor atmosphere and soil has been also used to characterize the spatial distribution and temporal variations of the gaseous concentration in background atmosphere, cave and soil air as well as the relationship between them.
Results reveal that the dynamics of ²²²Rn and CO₂ in the cave air may show different patterns depending on the time-frequency conditions and they are defined by the complex relationships between external and internal factors. Findings allow lately to preserve the quality of the cave indoors and to understand the possible risk derived from the presence of indoors hazardous substances.
Subesfera: una aproximación a la propuesta de un nuevo sistema ambiental en el Planeta
Pedro A. Robledo Ardila, Juan José Durán Valsero, Eulogio Pardo Igúzquiza, Raquel Morales García, Sergio Durán Laforet y Rebeca Álvarez Alonso
The term Subsphere (sub from the Latin below and σφαῖρα from the Greek sphere "globe"), is a proposal to define a new environmental system that is part of the Earth. It is constituted by a thin layer of irregular dimensions located in the terrestrial crust, but with a series of unique characteristics that make it different from the other environmental terrestrial subsystems. In addition, it is also the underground layer most accessible to humans and allows direct exploration and scientific research in the academic field of endokarst. The Earth is divided into different layers that define either mechanical systems or chemical systems, separating between deep layers (crust, mantle or core) or surface layers (lithosphere, atmosphere or hydrosphere). Among the surface layers, subdivisions into systems have been made to important differential characteristics, such as those defined in the atmosphere (troposphere, mesosphere, etc.) or in the hydrosphere (oceans, rivers, lakes, etc.). Within these large subdivisions, environmental systems have been classified, such as the biosphere or the cryosphere, among others. These systems have two major characteristics a) they are formed by sets of geosystems and/or biosystems and b) they are directly related with the large surface layers of the Earth that influence or are influenced by them; for example, with the atmosphere, the hydrosphere or their elements can influence the global climate of the Planet. The subsphere develops mainly in carbonate and evaporitic rocks, but also in detrital systems or rocks of other mineralogy. Its three-dimensional expression, from the surface of the Earth, reaches thousands of km3, millions of km2 of extension and a depth of tens of kilometers. It is an environmental system with very distinctive geological structures that clearly differentiate them from other portions of the lithosphere. Moreover, a very precise information could be obtained from sedimentary record as speleothems linked to processes operated millions of years ago on the Earth. One of the most spectacular expressions of the subsphere can be observed in large caves with great dimensions (Mammoth cave in USA 664 km; Sac Actum, Mexico 364 km, San Don Cave in Vietnam 40 million m3 or Krúbera-Voronya, Georgia 2.2 kilometers in depth). The subsphere has its own very diverse life system, with living organisms (macro and micro, animals and plants) that can only survive in this environmental system. The organization of life is based on a hierarchy with different levels of complexity, with smaller systems that are not always organized to form larger ones, contrary to what happens in the biosphere. In fact, micro life is grouped into ecosystems dependent on chemoautotrophic organisms due to the absence of light. Recent research has estimated that 70% of the Earth's microorganisms live underground. The subsphere also develops a particular atmosphere, with specific conditions of temperature, humidity and other chemical elements such as CO2, sulfur or methane among others. This allows a direct interrelation to occur with the large surface layers of the Earth, giving rise to the transfer of energy and chemical elements. These can be influencing the global climate system, either in the atmosphere, or in the oceans, seas or rivers, or the lithosphere itself (it can even interact with the deep layers of the Earth, hydrothermal systems) or with other subsystems such as the biosphere or the cryosphere. The connection between the subsphere and other environmental Earth systems can modify vital cycles such as the water or carbon cycles (it has been estimated that underground organisms represent more than 20,000 billion tons of carbon). In addition, processes in the subsphere can produce exchanges of matter and energy that can modify global dynamics. In addition, the aquifers are part of this proposal for a new environmental subsystem, as most of the fresh water available on the planet is underground, with an estimate made of almost 11 million km3 (half of the water contained in the cryosphere). Therefore, the subsphere is thought to be one of the large terrestrial environmental systems, expanding this small group of systems.
The importance of snow in the hydrogeology of a high relief karst system: Sierra de Tendeñera, in the Pyrenees mountain range (Huesca, Northern Spain)
Abstract
The Tendeñera mountain range, located in the Western Pyrenees in the province of Huesca (Spain), has a rectangular shape of 15 km in the E-W direction and 4 km in the N-S direction. Both the landscape, and a well-developed network of endokarst conduits have been conditioned by the structural disposition of highly dipping strata of a Mesozoic carbonate sequence that has been modeled by glacial and karstic processes. The highest parts of the mountain range reach an altitude around the snow line at 2600 m a.s.l. and with the maximum altitude of 2845 m a.s.l. that occurs at the Tendeñera Peak. The exokarst of these highest parts of the mountain range is populated by well-developed karst depressions (dolines) that connect with an extensive network of caves, with important vertical development, greater than one thousand meters. During winter the accumulation of snow in the highest parts implies the filling of dolines and potholes by snow and ice. Remains of ice and snow can stay until the end of august which implies several processes. First, snow and ice at the bottom of dolines and the entrances of potholes and caves behave as a secondary reservoir of water in a perched frozen aquifer. That implies that the baseflow during the summer is higher than it will be in absence of snow and the minimum discharge is reached in October rather than at the end of August. Secondly, the hydrochemistry of water discharged at the springs with a low content of dissolved chemicals and thus an unusually low electrical conductivity is very different to the one of springs in classical karst. All these aspects have been investigated in this work.
During a cave flood, sands and silts are transported through the cave system and a coating of this detritus is deposited on previous speleothem and cave surfaces. After the water recedes, speleothem deposition is re-initiated and detrital coatings get trapped inside the speleothems, while sedimentary sequences may be deposited in blind galleries or areas protected from the main water stream. This process can be repeated for centuries or millennia providing a unique and continuous (in the case of speleothem) record of paleofloods in a particular region. The closer speleothems and detrital sequences form next to the stream, the more often they record ordinary floods. On the contrary, those deposited in areas located far from the active flow or in the upper cave levels will only record extraordinary floods. Here we present the geomophosedimentary and chronological (U/Th, OSL) information derived from the Ojo de Val Junquera cave (Iberian Range, Spain) which hosts a great potential for paleoflood reconstructions. The cave comprises two sub-horizontal levels. The lower level corresponds with an ephemeral spring, only active during intense rainfalls, and includes six siphons. The upper level is at +11 m above the cave entrance and corresponds to a fossil cave level. Both levels are connected by shafts and ramps. That situation, together with the constrictions of the lower gallery, favors a water elevation up to 9 m during heavy rainfall events (>60 mm). Stalagmites from the upper level show single sand layers and brownish areas in relation to past low/high-frequency flood episodes respectively. U/Th ages show that speleothems grew during Marine Isotope Stages (MIS) 9, 8, 7, 6, 3, 2 and 1. In the lower level, near to the main cave outlet, a poorly consolidated detrital sequence formed by sands and gravels reaches the cave ceiling, indicating that the cave outlet was once blocked by these sediments. An OSL dating of this deposit reveals an age of 377±39 ka (MIS 10). Although the duration of blockage is difficult to determine, these non-consolidated sediments are easily removed if exposed to water flow. On the other hand, there is no evidence of successive blockages of this cave outlet or, at least, they are not preserved. Assuming that, we suggest that the cave morphology and the sediment arrangements were preserved through time and, subsequently, that floods recorded in the upper level represent extreme events. This hypothesis opens a window to study past floods in contrasted climate periods pre-dating the Holocene, although more chronological efforts must be carried out to date the rest of the detrital sequences. The dating of these detrital deposits can help to understand processes affecting the cave geometry and, thus, hydraulic head rises and help discerning between extreme and regular floods preserved in elevated cave areas.
“Famenne – Ardenne” Geopark was labelled by UNESCO in April 2018. Its area is about 1000 km² consisting of a SW-NE narrow band of about 15 km wide and 70 km long. Hercynian orogeny has tilted the Devonian formations to sub-vertical position leading to a contrasted lithological substratum. In the center of the area, a limestone strip ofsome 700 m wide and running east to west, is known as the Calestienne. Topographically, this is a scarp area between the higher Ardenne area, with a mainly sandstone substratum, to the south, and the shale substratum of the Famenne depression to the north. The valleys of three main rivers (the Lesse, the Lomme and the Ourthe) run through the Geopark, offering extraordinary geological potential that has provided much scope for the development of both tourism and scientific research.
It is along the limestone strip of the Calestienne that many (more than 1000) karst sites are inventoried in the Geopark. The biggest caves, i.e. Han-sur-Lesse, Rochefort and Hotton, are located respectively from west to east, inside the ~500 m-thick Givetian limestones, and consist in touristic infrastructures. The majority of the karstic phenomena (swallow holes, sink holes, travertine deposits, dry valleys, …) are observed where these limestones crop out.
However, the main impact of the geology on the landscape is observed where the Frasnian limestones of irregular shapes (reef lenses or narrow stratified formations) crop out and are interbedded with the Frasnian shale formations. Due to the frost wedging, landscape is subdivided in crests and depressions and finely serrated : crests locate the limestone formation and the depression the shale one. Using Lidar pictures, geological maps could be drawn just looking to local topography. Plunging anticline or syncline folds and faults are easily located to draw structural maps. Behind the topography, the land use is specific with the shaley depressions used for grazing, while the limestone hills are wooded and locally maintained to protect calcareous grasslands rich in characteristic flora and fauna.
The valorisation (conferences, visits, guide books, …) and the conservation of this exceptional landscape by the “Famenne-Ardenne Geopark” scientists is perfectly in line with the objectives of the Unesco Global Geoparks (UGG).
According to origin, karst caves are, of course, an underground phenomenon, but in the development of the karst they frequently appear as surface phenomena that we call “unroofed caves”. Due to the denudation, dissolution, and disintegration of carbonate rock and the subsequent lowering of the karst surface, which can be many dozen meters in a million years, underground karst caves are uncovered and with further dissolution can gradually disappear completely. Karstologists often explained various indentations in the karst surface recognized today as the lower parts of karst cave passages as various types of dolines, parts of former surface streams, or just the consequence of the lithological and tectonic characteristics of the rock. Here we show the frequency, significance, and characteristic forms of unroofed caves and place them correctly in the development scheme of the karst cycle. Many years of research of the Classical Karst in Slovenia have shown that unroofed caves are a relatively frequent surface karst form, certainly more frequent that we imagined before expressway construction works in Slovenia’s karst regions uncovered the karst surface. The first attempts to typify their characteristic forms and to explain their formation were born. Our results show the development from old caves to unroofed caves due to the lowering and dissection of the karst surface. Old caves are preserved by sediments, primarily alluvium and flowstone. We present forms of unroofed caves and determine their significance in studying the cavernosity of karst aquifers, the epikarst, and the surface. As a result of the latest research, we have become more attentive to these unique but characteristic karst surface forms. We summarize our many years’ experience of studying this interesting karst phenomenon and add the latest findings. We will focus on examples from the Classical Karst.
References:
Gams, I., 1965: On the Quaternary geomorphogenesis of the area among the karst poljes of Postojna, Planina and Cerknica.- Geografski vestnik 37, 60–101.
Knez, M., Slabe, T. (Eds.), 2016: Cave exploration in Slovenia: discovering over 350 new caves during motorway construction on classical karst, (Cave and karst systems of the world).- Cham: Springer, XIII, 324 p.
Knez M. et al. (Eds.), 2020: Karstology in the Classical Karst, (Advances in karst science).- Cham: Springer, XII, 222 p.
Karstification processes in coastal carbonate aquifers are strongly influenced by mixing of freshwater and seawater due to seawater intrusion (SWI) and periodic flow fluctuations on multiple time-scales. Here we use numerical modelling to investigate the impact of mixing mechanisms and long-term sea level fluctuations (regressions-transgressions) on the dissolution of calcite in coastal aquifers. Two-dimensional variable-density flow and transport simulations were performed considering different chemical compositions of the two end-members (freshwater and seawater) and sea-level fluctuations on scales of millennia. We analyze the reaction dynamics by focusing on the enhancement of porosity and permeability and the configuration of dissolution network patterns. We find that the resulting dissolution configurations are significantly influenced by the nature of sea level fluctuations. Thus, maximum dissolution rates occur, and zones of locally enhanced reactivity are created when the sea level rises from a global or local minimum level. Moreover, numerical results suggest that regular long period fluctuations induce the emergence of very large horizontal structures with enhanced porosity whereas irregular temporal variations induce maze dissolution network patterns. We also find that although high CO2 pressures in the freshwater induce higher dissolution rates, very similar patterns are obtained for the different end-members considered. Finally, our findings provide an alternative mechanism for deep karst formation associated to the Messinian salinity Crisis along the Mediterranean coast.
Sinkholes linked to cover evaporite karst in urban environments still represent a challenge in hazard and risk assessment. The Quinis hamlet, located in the mountain sector of Friuli Venezia Giulia region (NE Italy), in the Alta Val Tagliamento valley within the Enemonzo municipality, is heavily affected by sinkhole phenomena which deeply interested the infrastructures and houses present in the area. The area is characterized by a Carnian evaporitic bedrock constituted by gypsum and anhydrite mantled by alluvial and colluvial deposits. The climatic data recorded at the Enemonzo rain gauge station for the period 1995–2013 highlight a mean annual rainfall value not exceeding 2,500 mm. Rainfalls are abundant during autumn and springtime meanwhile during January and February they do not exceed a mean maximum value of 100 mm. In order to understand the evolution of these type of phenomena, a field-experiment started on the dissolution rate. In the area, in the 20 existing piezometers, at different depths, 50 evaporitic rock-samples (parallelepipeds having 8 cm of height and a square based of 4 cm) were exposed to the naturally occurring variation of relative humidity, air flow and hydrodynamics. The rock-samples were placed respectively in the aeration, in the fluctuation and in the phreatic section of the piezometric tubes. Data related to water level fluctuations, temperature and electrical conductivity were collected by using several data-logger devices. Every three months rock-samples were removed, weighted and the volume loss evaluated. The obtained results indicate that rock sample reduction is not only dependent to the water table fluctuations and to the number of days during which the samples are in contact with water but also to the location within the study area. Some of the rock-samples has been almost completely dissolved in three months, with dissolution rate values almost three times than expected if compared to the available literature data. This approach represents a novel contribution to the overall knowledge on karst and iper-karst processes where not only chemical dissolution occurs, but also the erosion processes can cause a loss of important volumes with noticeable impact on human-built construction.
Abstract
The Serrezuela de Carratraca Massif is a small two sq-km carbonate massif located near the villages of Ardales and Carratraca, Málaga province, in the South of Spain. It corresponds to an outcrop of a geological unit formed by Triassic dolomitic marbles and limestones, belonging to the Alpujárride Complex of the Betic Cordillera. The surface of this karstic massif lacks well-developed exokarstic features, however, there are several known caves, with significant dimensions, as well as a spring with relatively high flow rates. The most important caves are the Ardales Cave (1577 meters of horizontal development) and Sima Gorda (121 meters of vertical development). Other near-surface cavities have recently been identified in the slopes of the road that reaches the highest point of the Serrezuela. These small cavities have vertical tendencies and small diameter (less than 30 cm) and are interpreted as chimneys. Ardales Cave, located in the northwest corner of the massif, is a maze-type cave, with a great amount of cupolas inside, in the walls and ceilings of the cave. It is situated above the water table, without current hydrological activity. Sima Gorda is a vertical rising-shaft type cave, that reaches the actual water table and that contains different gases in its underground atmosphere (HS2, CO2 and methane). The main spring, located near Sima Gorda, in the Southern part of the massif, is the Baños de Carratraca spring (Carratraca Baths), a well-known sulphurous spring used for centuries as a spa. These characteristics, as well as the presence of large amounts of pyrite transformed into iron oxy-hydroxides (limonite) in the Triassic marbles, suggest the possible common origin of all of these phenomena, linked to a deep regional groundwater flow. The Serrezuela Massif is a region of deep vertical water discharge, that conditioned the hypogenic speleogenesis of the caves in the massif, induced by the mix of deep water and local recharge water. The tectonic structure has been an important factor in the past speleogenesis of the oldest caves (Ardales Cave), and currently determine the processes taking place in Sima Gorda and the surroundings of the Carratraca Baths. The presence of recent travertines near the spring suggest that the processes of dissolution-precipitation are still active.
The typical geomorphological forms of karst such as dolines or poljés can play an important role in surface runoff by constituting preferential infiltration points for rainwater directly towards the water table. Their identification is therefore important to evaluate the modalities of groundwater recharge, the orientation of the current and past drainage axes within the aquifers and also to evaluate the groundwater vulnerability. As such, many multi-criteria methodologies (Paprika, RISK, etc.) include this indicator of surface karstic forms presence in the in the vulnerability maps of water catchments.
Their identification and characterization (depth, asymmetry, nature of the bottom, etc.) is however delicate and tedious since it is often based on the field reconnaissance which is long and dependent on soil conditions (presence of vegetation or crops). The resulting mapping is rarely exhaustive. However, the democratization of LiDAR surveys offering high-resolution digital terrain models paves the way for semi-automatically, quickly and over a much larger area detection of these shapes.
On the territory of the Eaux-SCARS regional project (https://sigesaqi.brgm.fr/-Projet-Eaux-SCARS-.html), aiming to understand the functioning of the CArbonate aquifers of the north aquitain Basin in France and in particular the conditions of surface infiltration towards the aquifers. The DTM HR, resulting from Lidar surveys, called RGE Alti V2® and made freely available by the French National Institute of Geographic and Forest Information (IGN), could be used to establish a first inventory of the dolines.
A methodology has been developed to identify the hollows by calculating the contour lines at 1m intervals. These hollows are then qualified in order to distinguish those likely to be real dolines by using the Miller circularity index, their proximity to the building or the road network, the latter being able to cause dips that do not correspond to dolines. Beyond this identification, it is possible to characterize their average and maximum slope, their orientation or their geology.
This method has been compared with others (Cartannaz, 2015) and the results confronted to inventories and maps obtained from bibliography and field work. The results show the potential of the method to improve the knowledge of the study area karst geomorphology and characterize the vulnerability of aquifers.
Beyond the method, it is worth highlighting the use of the RGE Alti V2® offering a high-resolution DTM over the whole of France and allowing the method to be applied to departmental extensions at a lower cost and without the need to carry out an expensive and localized lidar survey.
Nearly 15,000 potential dolines have thus been identified over the 22,000 km² of Eaux-Scars territory and provide decision support map. Their validation must be continued in order to be able to use it in the future. However, the use of these data seems promising to improve the location of the various geomorphological forms, to obtain a large-scale inventory and to identify forms invisible to the naked eye in the field.
The use of water resources in karst environments has always played a preminent role in the choice of the sites for human settlements. Especially in arid or semi-arid climatIc conditions, water can be very rare at the surface, so that the underground becomes the place where to look for the necessary source of water and for establishing sites for religious and/or pagan cults as well. We present in this contribution three different sites from the Apulian karst of southern Italy, where water emerging in caves has historically represented remarkable sites for the local communities. The study areas covers the three main karst sub-regions of Apulia, namely the Gargano promontory to the North, the Murge plateau in central Apulia, and the Salento peninsula farther South.
The first cave is Scaloria, in the Gargano promontory, where testimonies of the cult of water during Neolithic times have been documented by archeological sources. Locating vases and similar containers in sites close to cave lakes or dripping places was interpreted as linked to cult of water rather than to the need to provide with drinkable water, given the often difficult way to reach such sites.
Analogously, at Minervino Murge, in the High Murge plateau, the cave dedicated to Saint Michael Archangel hosts places to collect the water, dripping down from both the cave ceiling and speleothems. Being one of the many caves in Italy dedicated to the cult of Saint Michael Archangel, the site has played an important role in history and in the pilgrimages along the main religious routes of southern Italy.
Eventually, at Manduria in the Taranto province, Fonte Pliniano (Plinian Source) is a spring that was originally documented by Pliny the Elder in his work Naturalis Historia. The spring opens at the contact between the Cretaceous limestone bedrock and the overlying Upper Pliocene – Lower Pleistocene calcarenite deposits. Within these latter, the presence of marly-clay interlayers is at the origin of the formation of perched water tables, and as a consequence water outlets in the forms of diffuse springs are observed. Given the archeological relevance of Mandyrion (the old Manduria), one of the most important Messapian towns (Messapi is the name of the indigenous population living in southern Apulia before the Romans), the spring has always been a remarkable site for the area, also as a place of cult.
Through illustrations of the three cases above, our goal in this work is to highlight the strong link existing between water and human presence, with the related cults, in karst settings.
Cave sediments are valuable archives of Quaternary landscape and environmental change in karst regions. In northern France, recent research revealed the development of four regional cave levels formed by the incision of the River Seine during the past one million years. The lowest cave level, perched 15 m above the present river, corresponds to caves intercepted by the Caumont medieval underground quarry near Rouen. Here, the excavation of a network of sediment filled caves for exploration purposes provided many excellent sections for study. This work aims to reconstruct the history of sediment infilling in the Caumont chalk caves and link this to the evolution of the Seine valley during the Pleistocene. The ongoing work includes: i) cave surveying to reconstruct the 3D geometry of the cave conduits intercepted by the underground quarry, ii) study of 15 stratigraphical sections to define depositional sequences, iii) micromophological observations to characterise the sediment and their sources, and iv) U-Th dating of speleothems to provide the chronological framework. From the data we present a preliminary evolution model of Caumont karst system and link it to the fluvial dynamics of the River Seine. The model involves five phases: (1) infilling of cave conduits during glacial periods prior to 127 ka, (2) speleothem precipitation at around 127 ka, (3) partial erosion of the cave infill during subsequent stages, (4) renewed sedimentation derived from the erosion of surficial formations (e.g. Pliocene sand) prior to 7-10 ka, and (5) clay decantation and shelfstone (speleothem) precipitation in pools during Holocene.
A novel method to locate karst cavities underground using electrical resistivity tomography and the posterior population expansion inverse method.
Abstract
The traditional invers methods used to interpret electrical resistivity tomography (ERT) measurements are not able to properly characterize karst environments due to the strong contrast between air or water filled conduits and the surrounding ma-trix. In fact, these inversion methods were originally conceived to interpret rather uniform lithologies. The geostatistical Posterior Population Expansion (PoPEx) method is an inversion tool designed to deal with data presenting variations such as those encountered in a karst system. The advantage of this method is that it directly returns maps indicating the probabilities of encountering conduits instead of the usual deterministic set of resistivity values that one has to interpret manually. The method is applied here to invert synthetic 2D ERT data of karst systems. At this stage, PoPEx is computationally demanding and needs to be standardized. However, it has proven to be effective for simple and synthetic cases and so, it is though relevant to further investigate its wider application.
The Las Loras Geopark is located in the contact between the southernmost sector of the Basque-Cantabrian Arc and the detrital sediments of the Tertiary of the Duero basin. The entirety of its territory is made up of carbonate aquifer materials that constitute permeable formations. Hydrogeologically, these formations are shared by the Ebro basins (Páramo de Sedano y Lora groundwater body, ES091MBST002) and the Duero basin (Quintanilla-Peñahoradada-Las Loras groundwater body ES020MSBT000400004), being of the order of 80% of the geopark drained towards the latter. The hydrogeology of this region is out of date. Previous works present an outdated description of the aquifers present in this geopark. This work presents a state of the art of regional hydrogeological knowledge at the geopark scale and identifies the main lines of work for a detailed hydrogeological characterization. The work methodology includes a bibliographic review of the materials published in various sources, both official and sources from other non-profit entities that have contributed to regional and local knowledge in the region. The capture of data from official sources, mainly Duero River Basin Authority (CHD) and the IGME-CSIC database, and their treatment allow obtaining a snapshot of the state of the aquifers on a regional scale. The fundamental results lead to the need for a hydrogeological study based on new field data that allow the following objectives: to identify the priority directions of the groundwater flows, to locate the groundwater divide, to locate the sinks, water courses (permanent and temporary) and springs, as well as other elements of hydrogeological interest that not due to their minor presence, are of less interest (wells, caves, galleries), some of which have already been identified as Sites of Geological Interest, and are compiled on the page geopark website. Others, potential new points are commented in this work. This results are part of the IGCP-730 project funded by UNESCO's Geological Sciences Program.
With an area of 36 square kilometres Cerkniško Polje is the largest karst polje in Slovenia, which is flooded up to 8 months per year. It has more than 300 years history of scientific exploration, much of which has focussed on studying of the flood dynamics of the system and the resultant adaptation required by the local communities. As a result many construction activities have been carried out in the past in order to regulate the flooding levels.
The aim of present multidisciplinary ecohydrological study of the polje is to relate the flooding dynamics with ecological habitat types and land use. The study comprises regular field measurements of hydrological parameters (temperature, conductivity, pH, dissolved oxygen, etc.) at up to 40 locations at springs, streams, swallow holes and water active caves. Additionally, periodic discharge measurements and sampling for the water quality monitoring are taking place. Five autonomous instruments for continuous measurements of water level, temperature and conductivity are installed at swallow holes, estavelles and water active caves. The studied area also has records available for more than 60 years of continuous hydrological and meteorological measurements, as well as detailed maps of habitat types from 2009.
Analysis of the long-term water level measurements has revealed evident changes in the flood dynamics over the past decades, reflected in shorter and less extensive (with respect to both area and volume) floods. According to the visual inspection of long-term measurements this shift is most likely linked to both climate change (distribution of the rainfall) and human construction activities (regulating of river beds, widening of swallow holes, construction of dams, etc.). Different flooding patterns clearly results in the different spatial distribution of vegetation habitats across the polje, as well as taking into account land use changes, such as intensity, frequency and extent of harvesting. Even though Cerkniško Polje is noted for its rich biodiversity, the changeable flooding pattern already appears to be impacting on the different composition of plant species as well as anecdotal disturbances in fish and bird life cycle.
Acknowledgement: This study is taking place within the project “Evaluating the ecohydrological dynamics of the Cerkniško Polje intermittent lake using a multidisciplinary approach”, no. Z6-2667 and Karst Research programme (P6-0119), financially supported by the Slovenian Research Agency. Equipment has been supported by projects “Development of research infrastructure for the international competitiveness of the Slovenian RRI space – RI-SI-EPOS and RI-SI-LifeWatch.”
Karst vadose zone is a network of channels, voids and fissures in which mass and heat transport is controlled by a combination of gravity-driven water flow and airflow. Airflow in karst vadose zone is determined by the structure of the channel network and its connections to the outer atmosphere. A dominant factor in subsurface airflow is the chimney effect, where pressure differences that drive the airflow are caused by density differences between internal and external air. Airflow can also be caused by external barometric variations or by external winds causing near-surface pressure variations. Using long-term observations in several caves in the Dinaric Karst, we show how the different driving forces of subsurface airflow interact and how the resulting airflow patterns are reflected in typically measured parameters such as temperature and pCO2. Most of the results presented are based on observations in a remote "dead-end" passage in Postojna Cave (Slovenia), where extreme temporal and spatial variations in CO2 have been recorded. We explore the mechanisms behind these fluctuations and show their impact on the chemistry of percolating water. Finally, we introduce and use a new numerical model to investigate how the diversity of airflow patterns caused by the chimney effect is related to the geometry of a simple cave system with two entrances.
The assessment of carbon cycle in the Earth-climate system is one the highest challenge in science nowadays. It still remains some key knowledge gaps and uncertainties concerning the budgets of greenhouse gases (GHG) at ecosystem scale. Covering up to 25 % of the land surface and acting alternately as CO2 source or sink, karstic subterranean ecosystems play a decisive role in carbon cycle in terms of it contribution to global balance of this GHG. The interactions between geological, microbiological and chemical processes are responsible for the physical-chemical properties of the atmosphere and especially for changes in its composition. However, there are still essential gaps in our knowledge about the possible feedback mechanisms between the environmental-microclimatic conditions and the rates and type of activity of microbial communities in natural subterranean ecosystems.
Here we study, for the first time, the interactions between the microbiota and the subterranean ecosystem (Pindal cave, northern Spain), with a special focus on environmental controls and feedback, as a key challenge to clarify the effective and accurate contribution of subterranean ecosystems to the global carbon cycle. We applied in situ and real-time monitoring diffusive fluxes using closed chamber-based gas exchange system coupled with NDIR and FTIR gas analysers, a coeval δ13C geochemical tracing by cavity ring-down spectroscopy and metagenomics analyses, to evaluate and quantify the CO2 fluxes from microbial communities associated with cave sediments and bioinduced-carbonate deposits as moonmilk.
Our results on carbon isotopes and CO2 fluxes suggest a source of organic carbon due to Actinobacteria metabolism linked to calcite moonmilk formation and the higher CO2 fixation by the chemoautotrophic bacteria, while in the raw cave sediments (i.e. those not covered with moonmilk deposits) prevails the carbon oxidation of this organic matter by autotrophic and heterotrophic communities, leading to an in-situ production of isotopically heavier CO2. The net CO2-flux rates from the raw cave sediments to cave air varied from 24 to 27 mg CO2 per hour, twice higher than the maximum flux rates registered on the moonmilk surfaces. These findings demonstrate how microbiomes influence the systems in which they inhabit, having direct implications on current challenges of the scientific community including, the study of potential extra-terrestrial analogues through biosignature detection and the design of appropriate strategies for the conservation of subterranean sites with a valuable heritage.
Karst aquifers and, especially, their caves and springs, are singular sites of the geoheritage due to their scientific and touristic interest. In fact, both of them are recognized as habitats of special protection for the Habitat Directive (92/43/CEE) and would be used for local development in UNESCO Global Geoparks (UGGp). However, their management continues to be a challenge at the present day since karst aquifers, and their associated springs and caves are vulnerable environments used by many actors (e.g., speleologists, local people, tourism companies, scientists) and affected by land uses and changes in surface hydrology and groundwater. The work aims the implementation of a suitable management of karst springs and caves in the Courel Mountains UGGp in the framework of the project IGCP-730, titled Hydrogeological significance of Mediterranean geoparks, and founded by the International Geosciences Programme of UNESCO. The management includes (1) the elaboration of the hydrogeological map, karst cave and spring inventories of the UGGp, (2) the definition of geosites related to the karst, (3) water point inventory, (4) cave monitoring, (5) dissemination of cave geology and karst hydrogeology via publications, activities and visitor centres, and (5) geoconservation actions. The hydrogeological map of the UGGp was performed in a Geographic Information System (GIS), which includes also 70 springs (5-250 l·s-1 discharge), wells, cave entrances and cave maps. Eight of these features were defined as geosites of the Courel Mountains UGGp according to their nature, scientific and use values. The hydrogeological characterization focused on terrains which are recognized as aquitards within the metamorphic bedrock (81% of the UGGp extension), porous aquifers related to alluvial and slope deposits (12%), karst aquifers (5%) and aquitards associated with glacial and other deposits (2%). This characterization has recently started with a hydrochemical study of surface and groundwaters based on 24 water samples collected in a field campaign last September-2021. The physico-chemical analysis obtained provided the pH, electrical conductivity, oxidability and the anions and cation concentrations via ion chromatography, which allowed a first identification of the dominant hydrochemical facies. The goal of this hydrogeological characterization is also addressed to stablish an environmental monitoring network of the environmental conditions of the UGGp for the future and the study of its resilience to global change. The water quality is one of the key indicators for the good state of terrestrial and aquatic related ecosystems. Cave monitoring comprises the short surveying of CO2 and temperature of three caves and it comparison with the visitor affluence, revealing minor impacts. The singularity, functioning and relevance of karst and caves aquifers were promoted for all audiences by means of publications, conferences, fieldtrips, and a visitor centre focused on karst caves and groundwater. Finally, geoconservation actions involve volunteers (speleologists, local people, teenagers) for cleaning and removing of waste accumulations deposited in cave entrances. Altogether, the initiatives allow us to manage karst caves and springs in a holistic and pragmatic way based on the scientific knowledge, and involving local administrations, cave users and native people.
The growth of the urban slick represents one of the most significant challenges to territorial planning in karst territories. This growth undermines the city's planning systems, generating irregular settlements, disarticulated human settlements, lack of services, coupled with ignorance of ecosystems, a situation that is typical of cities in the northeast of the Yucatan peninsula. The absence of a knowledge system of the physiographic characteristics of the territory has overlooked the karst characteristics of this region, adapting the buildings to a highly complex urban system. This work aims to develop a model for the recognition of karst depressions and its assessment based on the use of urban land in order to guide karst action policies in the city. The study was carried out in the city of Playa del Carmen, Mexico. For this, the cartographic bases were developed in order to recognize the depression units in the city, which were derived from data from seven INEGI LiDAR datasets, with a resolution of 5 m in the horizontal axis, constructing a mosaic dataset that covers the urban stain of the city. Through a management of the original LAS data, an altitude contour map (from 0 to 43 meters) was modeled with an equidistance of 0.25 cm in the vertical axis and with polygons of 25m2 of resolution; subsequently, the positive and negative level contours (depressions) were differentiated. Which were characterized and differentiated by their origin (natural or anthropogenic), type (sinkhole, uvala or polje), depth (from 0.25 to 40.5 m), land use and population density at the city block level. Results: 1350 depressions were recognized, of which 59% are orbed and orbed to oval sinkholes with less than 1 m depth. Four sectors of the city are distinguished with medium and high housing density, high population density and high density of depressions (between 50 and 90%). The identification of the units is based on the principles of remote perception and geographic information systems, so that some spatial units will be over-estimated in their number or underestimated in their territorial context of the urban area of the city of Playa del Carmen. Most of the sinkholes and uvalas are located in private and business properties (mainly nurseries, ornamental gardens and water vending), government (in the form of theme parks, within land and public parks) and in vacant lots (most are characterized by being open-air dumps) and about 30% is located in the road system. The present analysis model can be implemented in all the northern cities of Quintana Roo under an ecosystem services scheme of these karst units.
Ice melting is one of the most common indicators while showing consequences of increasing air temperatures and changing precipitation patterns. The most notable reports of its shrinking are known from polar regions and mountain glaciers, however significant loos has also been observed in karst (ice) caves.
Among 14,169 registered caves in Slovenia there are 230 with perennial ice. We present observations of cave microclimate and ice volume changes in 10 ice caves. Initial visual estimation of ice volume changes were followed by continuous air temperature measurements and topographic measurements using tape meter and fixed points. Recently, measurements are upgraded by terrestrial LIDAR scanning. This method enables detailed measurements of the cave geometry, ice surface and detection of changes of the position of individual objects (rocks). For this purpose, two ice caves with relative easy access (Velika Ledena Jama v Paradani and Snežna Jama na Planini Arto) were chosen for more detailed observations. Scanning is performed twice a year – in the spring and in the autumn. First analyses showed significant loss (up to 50 cm) during summer melting period and weak ice accumulation during winters.
The preliminary results encourage further measurements to improve understanding of seasonal and long-term ice dynamics and its relation to climate changes. Periodic sampling of ice and analyses of its chemical and isotopic composition will give additional information about characteristics and past development of ice.
Acknowledgement: This study is taking place also within the Karst Research programme (P6-0119), financially supported by the Slovenian Research Agency. Equipment has been supported by projects “Development of research infrastructure for the international competitiveness of the Slovenian RRI space – RI-SI-EPOS and RI-SI-LifeWatch.” The operation is co-financed by the Republic of Slovenia, the Ministry of Education, Science and Sport and the European Union from the European Regional Development Fund.
The morphological study of the relief allows an adequate classification of the form, elements and structure of the karst relief that helps in the determination of the genesis, evolution, and geomorphological dynamics. Currently, there are applied studies on depressions and geomorphology in the Yucatán Peninsula. The implementation of tools with new remote sensing and GIS methods and technologies helps to develop fine and more accurate studies on relief models with low altitudinal contrast. The present study aims to analyze and classify karst depressions in the north-eastern coastal region of the Yucatán Peninsula, Mexico. The karst relief was analyzed based on the construction of a Digital Elevation Model (DEM) obtained from two sources of altimetric data: a) construction of a mosaic of LiDAR (light detection and ranging) data, and b) use of data obtained from Aster GDEM (Aster Global Digital Elevation Model) stereoscopic satellite images, which were coupled to generate altimetric data and morphometric base models. Subsequently, the contour lines were generated, and the depressions were differentiated, recognizing the escarpment and bottom. With the above, the data were processed and classified through analysis of the following variables: area, perimeter, relative depth, volume, azimuth of the X axis (major axis), classification by morphology of the profile in plan (V-shaped, U-shaped and ∐ (box) shaped.) and classification by elongation index (sinkholes, uvala-sinkholes, uvala, and poljes). We identified 10,672 depressions, taking as criteria a minimum depth of 1 m; the major axis ranges from 22 to 835 m, with an average of 128 m; the maximum depth is 45 m and the minimum 1 m, with an average of 3 m. By means of the profile typology, it is recognized that 67% correspond to V-type depressions, 23% to U-type and 10% to box-type profiles; the latter have a bottom in contact with the phreatic mantle. With respect to the elongation index, 18% are sinkholes, 42% are in transition from sinkholes to uvala, 13% are uvala and 27% are polje, being the uvala and polje those with a complex configuration in their shape, which is due to their topographic position close to the coast and the structural control of the Holbox and Tikul fault system, with a main orientation of NE 7°SW and secondary SE83°NW, showing the disjunctive influence in an orthogonal system. In addition, an inventory of the coastal depressions of the north-eastern Yucatán Peninsula is presented as a basis for the management of the coastal karst territory at the municipal level. The differentiation of the units by their typology and their characterization is an input for the management of the units at the local level. The limitation of the study focuses on the resolution and the coupling of the source data model, the LiDAR data have a data resolution of 5 m for the X, Y axes, and with 0.50 cm for the Z axis; the Aster GDEM data have a resolution of 30 m for the X, Y axes, and with 1 m for the Z axis, however, the coupling of models serves to complement the inventories of those territories where LiDAR data do not exist in the Yucatán peninsula.
Unlike surface water reservoirs, that can be easily quantified and monitored, underground conduits in karst systems are often inaccessible and therefore challenging to monitor. Our knowledge of inaccessible karst aquifers is therefore mostly based on tracer tests and on the analysis of karst spring hydrographs and chemographs. As a complement to these methods, seismic noise analysis was suggested to monitor ground water storage in a fractured rock aquifer (Lecocq et al. 2017). In underground karstic environments, seismic noise monitoring was able to detect hydrological cycles and monitor the groundwater-content variations (Almagro Vidal et al. 2021).
This study is a part of the SISMEAUCLIM project which aims to image the flow of water in 3 dimensions over time from rainfall to the outlet in the Fourbanne karst aquifer (Jura Mountains, Eastern France, Jurassic Karst observatory). Our approach relies on passive ambient noise methods that will be coupled with hydrological parameters in order to monitor underground water flows. The underground conduit is accessible through a drilled shaft and instrumented by two 3-component seismological stations and a water-level probe.
We’ve performed a joint analysis between the seismic noise spectrograms and the underground river water level in order to quantify the increase in noise intensity as a function of water level and characterize the noise source. Then we plotted the noise amplitude in function of the water level revealing a hysteresis due to the transport of gravels during flooding and their deposit during recession as reported elsewhere by Burtin et al. (2008) for surface rivers. As a second step, we applied a new approach based on the machine learning random forest (RF) algorithm and continuous seismic records (citation), to find characteristic signals to predict the underground river water level. The method consists on the computation on a sliding window of seismic signal features (waveform, spectral and spectrogram features). The first results indicate that the RF algorithm is capable of accurately detecting flooding periods and reproduce the underground hydrology the corresponding flood, with a mean absolute error not exceeding 5%.
This work shows that underground hydrological processes occurring in the aquifer, such as water flow or sediment transport, can be precisely monitored using the spectral analysis method. In addition, the RF results are a first promising outcome for the remote study of water circulation in karst aquifers using seismic noise.
The Orbieu drains a catchment area of 680 km² in the Aude department (south of France).
Sometimes, the river, experiences deadly floods that can reach a flow of 2000 m3/sec.
Some of them, unforeseen, testify to a lack of knowledge of the impact and behavior of the karstic component of the basin on hydrology.
Located at the crossroads of several influences, the Orbieu karst develops itself in a multitude of geological and climatic contexts, with however a capacitive structural constant.
At the request of SMMAR11, thanks to a network of nine multiparameter monitoring stations, installed in the key hydrogeological sites of the karstic units, we covered 3 hydrological recording cycles. The choice of live remote transmission of data allows optimal targeting of gaugings for qualitative calibration, and dying tests.
The results obtained show an adaptation of high water underground drainage in response to the climatic specificities of the watershed.
There is a dichotomy between rainfall impulse and increase in flow of the Orbieu. It seems to be explained by the existence in the aquifer of pre-saturation thresholds that the study highlights for each unit. These thresholds condition the “buffer” role of the karst, or, conversely, its sudden contribution to surface floods by phasing (activation of overflows). It becomes abnormally high by concentration when a refusal to infiltration occurs (rain of constant but prolonged intensity).
In the context of climate change, the conclusions of this pilot project initiate reflection on “karstic alert thresholds” in other limestone regions with high meteorological risk.
Time series of hydrochemical parameters can support the investigation of the subsurface functioning of karst systems as well as the identification of dominant hydrological processes and conceptual model structures. Nevertheless, high costs of sample collection and analysis cause hydrochemical data to be rarely available at a sufficiently high temporal resolution to understand the transport processes of an investigated system. The electrical conductivity (EC), however, can be cheaply and continuously be measured by an EC sensor and represents an integrated signal of major ion concentrations that can be obtained at e.g. a karst spring.
Here, we present a method to derive high frequency individual ion concentration time series and their uncertainty from EC time series measured at specific karst springs. From continuous electrical conductivity and low frequency ionic measurements, the method aims to obtain major ion concentrations at the same temporal resolution as the EC time series. Due to the large ion concentrations and complex speciation characterizing the discharge of karst springs, the concentrations of each element occurring as free ions and as aqueous complexes were computed separately. The concentrations of the elements occurring as free ions were computed based on their contributions to the total electrical conductivity, whereas the concentrations of the elements as part of aqueous complexes were obtained with speciation calculations. In order to investigate how different degrees of karstification affect the data requirements and applicability of the approach, the methodology was tested in two different springsheds, i.e., the Kerschbaum spring in Austria and the Baget spring in France. Both springsheds show similar catchment size and land cover characteristics but different geological and climate characteristics. To be precise, the Kerschbaum spring originates from a dolomitic karst system that is mainly fed by diffusive recharge, while the Baget spring represents the outlet of a limestone system dominated by concentrated recharge.
Our results highlight that the EC signal of karst springs can be used to obtain major ion concentrations at high temporal resolutions. In the prediction of the temporal dynamics of major ions, we identified the lowest uncertainties for those ions with a large contribution to the total EC signal and with a low relative variability, i.e., Ca2+, Mg2+ and HCO3-. The results show that complexation processes are significant for SO42-, Ca2+, Mg2+ and HCO3-, whose neglection would lead to an underestimation of the total concentrations.
The proposed methodology allows us to obtain high-resolution ion concentrations time series, without performing continuous ionic measurements, and can thus support future works on karst systems conceptualization. The investigation of the proper temporal resolution, required to apply this method to different karst formations, can further guide fieldworks and sampling campaigns.
Snow recharge is an important dominant hydrological process in the high altitude mountainous karstic aquifer systems. In general, widely used karst hydrological models (e.g., KarstMod, Varkarst) do not include a snow routine in the model structure to avoid increasing the number of model parameters while representing the complex hydrological process. As a result, some processes are not represented well, which questions the optimality of the results that can be obtained under available datasets. To overcome these drawbacks, the SCA routine is developed and applied on the snow dominated karst catchments as a novel pre-processing method. The SCA routine is driven by the temperature, precipitation, and satellite-based snow observation datasets while classifying the precipitation input into three physical phases (rain, snow, and mixed) based on the temperature datasets to distribute each phase over the catchment using satellite-driven Snow-Covered Area (SCA) products. In this study, we created SCA time series by using the MODIS daily snow cover products, and we used SPOT images, and ground observations for validation of the SCA time series. To test the efficiency of the SCA routine we used KarstMod model to simulate the karst spring discharge in a well-observed snow-dominated karstic aquifer, Hochschwab catchment, Austria.
The Blautopf, Germany’s second largest karst spring, offers a unique opportunity to study the water quality year-round and allows direct access to groundwater/sub-surface streams by two cave systems, Blauhöhle (Blue Cave) and Hessenhau Cave. The 165 km² catchment can be subdivided in two sections of similar size based on the karst drainage. In both sub-catchments the aquifer can directly be accessed by caves that are connected only shortly before the water arrives at the spring. Each sub-catchment has a different land-use, which gives an excellent possibility to study respective water quality impacts both in spatial resolution in the cave system and in temporal resolution at the spring. It further provides a unique opportunity to study transport processes of particles in a greater detail.
The Blautopf spring was intensively monitored following an evenly distributed, moderate rainfall event with a total of 40-55 mm precipitation over a duration of four days as well as during the snowmelt period. In-situ parameters, major ions, trace elements, rare earth elements (REE), total organic carbon (TOC), particle size distribution (PSD) and turbidity were measured in short intervals and all parameters showed a moderate response upon the arrival of freshly infiltrated surface water. A particular focus was put on the monitoring of coliforms and E. coli bacteria, which exhibited an up to 100-fold increase in bacteria numbers after approximately three days following the rainfall event.
Additionally, selected water samples were collected in the cave streams before, during and after the rainfall event. The caves are very difficult to access, thus only a very limited number of water samples could be taken. These samples give a good prediction about the change in water quality several days ahead of the arrival at the spring. The significant differences in the water quality of the two sub-catchments support the paired-catchment approach inside the karst and cave system.
The here presented novel paired-catchment approach allows a better understanding of the spatiotemporal transport behavior of solutes, organic matter, suspended particles and fecal bacteria in karst systems in the background of different land-use activities.
In Switzerland, karst environment covers around 20% of the total surface and contributes just as much to drinking water supply. Yet, karst systems are extremely sensitive to surface pollution because of their high hydraulic conductivity. Karst network simulations can help protecting water quality, but are currently time-consuming and complex to operate.
Here we present pyKasso, a python3 open-source package intended to easily and quickly simulate karst network using a geological model, hydrogeological, and structural data. It relies on a pseudo-genetic methodology where stochastic data and fast-marching methods are combined to perform thousands of simulation rapidly. The method is based on the stochastic karst simulator developped by Borghi et al (2012). It is extended to account for anisotropy allowing to simplify the algorithm while accounting better for the geological structure following the method presented in Fandel et al. (2021). Statistical geometrical and tolopolgical metrics are computed on the simulated networks and compared with the same statistics computed on real karst network in order to evaluate the plausibility of the simulations.
Geological model, hydrogeological, and structural data are given by the user when stochasticity of the karst generator is controlled by 1) a configurable random fractures generator, where density, orientation, dip, and length can be set for each fracture family and 2) a configurable random inlets and outlets sampler, which distributes the order of karst network starting points during conduit generation. During the execution, pyKasso allows the user to easily modify the settings on a model given high flexibility to test features and to perform sensibility analysis. Fast-marching methods are used to quickly calculate time travel maps to determine a quicker path for water droplets where karst conduit could appear. Those maps are derived from a field of constraints, which are conceptually equivalent to water velocity. According to the situation, the user can perform the fast-marching method with an anisotropic algorithm, allowing to get more realistic field constraints. Gradient descent calculated on the travel time maps allow to obtain stochastic conduit karst networks. Finally, performing several times this procedure allow user to calculate the mean of all the previous computed karst network and give a probabilistic map of the presence of a karst conduit.
The present package has been developed in order to support 2D and 3D cases and has been applied to realistic study cases such as the Tsanfleuron karstic system (CH) and the Gottesacker karst aquifer (AUT). The proposed work aims to review the main steps of the modeling approach, to illustrate the different features and to presents some synthetic examples.
REFERENCES
Fandel, C., Miville, F., Ferré, T., Goldscheider, N., Renard, P. 2021: The stochastic simulation of karst conduit network geometries using anisotropic fast marching, and its application to a geologically complex alpine karst system. Submitted to Hydrogeology Journal
Borghi, A., Renard, P., Jenni, S. 2012: A pseudo-genetic stochastic model to generate karstic networks, Journal of Hydrology, 414–415, https://doi.org/10.1016/j.jhydrol.2011.11.032.
A novel approach has been developed for the prediction of spring discharge time series based on regional climate model projections. For this purpose, a combined stochastic-analytical modelling approach was developed. This approach integrates the advantages of traditional stochastic methods such as regression analysis with physics-based analytical baseflow models.
During the stochastic modelling of karstic spring discharge, hydrograph peaks usually show a reasonable correlation with daily rainfall data, but spring baseflow has insufficient correlation with rainfall. This is a direct consequence of the physical functioning and the dual hydraulic behaviour of karst. While hydrograph peaks (flood) originate from direct recharge into the aquifer, baseflow originates from the release of water infiltrated into and stored in the low permeability matrix blocks. Baseflow is temporally delayed compared to rainfall and discharge peaks. For this reason, it is not possible to adequately describe these to different physical processes with one regression function. While flood discharge peaks can be approximated through the application of regression functions between rainfall and discharge, the description of baseflow discharge requires the application of physics-based analytical functions.
The combined stochastic-analytical modelling method involved the establishment of regression functions between rainfall and discharge peaks, and the simulation of flood discharge based on future rainfall time series predicted by regional climate models.
The baseflow component of spring discharge was simulated using 2D analytical solutions for baseflow discharge. Composite exponential formulae were applied for describing baseflow, where fitting parameters were calibrated based on historical rainfall and discharge data.
Prediction of future spring discharge time series comprised the following steps:
1. Selection of climate projections assumed to describe future climate conditions.
2. Setting up regression models between rainfall and peak discharge.
3. Comparison between measured and simulated discharge time series to characterize model efficiency.
4. Simulation of future peak discharges based on selected regression functions.
5. Establishment of Master Recession Curves (MRC) for describing the baseflow component of spring hydrographs.
6. Selection and calibration of analytical model describing baseflow discharge.
7. Simulation of future discharge time series based on the combined stochastic-analytical method.
8. Analysis and descriptive statistics of predictive model results.
A combined stochastic-analytical modelling of discharge hydrographs at the Bukovica and Bijela springs of the Piva River catchment, Durmitor area, Montenegro was undertaken. The applied climate model projections originate from the EURO-CORDEX ensemble. We investigated the Representative Concentration Pathway scenarios RCP4.5 and RCP8.5.
Predictive modelling was undertaken for the periods 2006-2010, 2021-2050, 2071-2100.
Regression functions between rainfall and discharge were established using historical data collected between 2008-2010. Baseflow recession was described using a double-component exponential model, where parameter fitting was performed on Master Recession Curves. Rainfall time series from six selected Regional Climate Model scenarios were applied to predict discharge.
The applied method is suitable for simulating the temporal distribution and magnitude of discharge peaks and also for baseflow recession. The applied model scenarios predicted the probability and magnitude of discharge which were represented through Flow Duration Curves.
Wherever underground work is performed in a limestone unit that is prone to karstification, an assessment of the risk due to the possible existence of karstic void should be done during the project planning phase. Karstified rock mass are characterized by a semi-permeable matrix that is generally drained by a set of karst conduits that can conduce large quantities of water. These conduit diameters can range from some centimetres to a few meters, depending on many factors including the lithological rock properties and the size of the catchment area. Depending on the position relative to the groundwater table the cave voids can be submerged, only temporary water bearing. Especially submerged kart conduits can be very problematic during tunnel construction. Consequently, estimations of possible water inflow are capital to prepare remediation procedures. Since they are affected by a wide range of site dependant conditions, it can be difficult to robustly quantify them. The aim of this work is to provide a methodology for probabilistic estimation of water inflow from karstic conduits during tunnel construction. Possible realistic stochastic karstic networks are first generated based on fracture and stratigraphic information. Secondly, a discrete–continuum modeling approach is applied to the networks, allowing the flow calculation to take into account the exchange between the matrix and the conduits, the influence on the feeding of the network from other aquifers, and to characterize the type of flow regime at each point in the model (turbulent or laminar). Different hydrogeological scenarios are tested for the estimation of the long-term possible inflow, such as different pressure gradient or different matrix permeability values. The output of a few tausend stochastic flow returns a probability distribution of potential inflow, depending on the configuration of the system. It shows that even if the most important factors controlling the long-term inflow are the contact conduit aperture diameter and the pressure gradient in the system, the general network shape and connectivity can significantly affect the potential discharge in a tunnel. This study demonstrates the advantages of using such stochastic workflow in the project planning phase of underground works in karstic environments. Using such methods could greatly reduce uncertainty and consequent increase the safety for tunnel workers and reduce unplanned project delays and cost.
Carbonate aquifers are known as a major source for drinking water in the Mediterranean regions and in other part of the world. Therefore, qualitative and quantitative estimation of the groundwater resource are crucial, especially in area with densely populated areas. Stable water isotopes of surface and ground waters can be used to study water mixing, define recharge area, or identify fast infiltration in karst areas. It relies on (1) the variability of the input signal over space and time at a catchment scale, related to the rainfall isotopic signature, and (2) the flow, storage and mixing within the karst system, related to the soil – epikarst – unsaturated zone – saturated zone hydrodynamic behaviour. Mean groundwater isotopic signature can then be different from the mean total rainfall signature, depending on the effective rainfall amount.
The case study, located in south-eastern France, offers an attractive hydrogeological context to explore the seasonal variability of the stable water isotopes of rainfall and effective rainfall at a catchment scale. The 500 km² area studied is characterized by a large range of elevation, from the sea level to 1 148m, giving an expected significant contrast of rainfall isotopic signature. Monthly water samples have been collected in-situ during two years on groundwater (boreholes, karstic springs) and surface water (rivers), which are compared to rainwater (monthly sampling at two distinct elevations and distance from the Mediterranean Sea) to constrain the isotopic signal of the recharge.
The isotopic signature of the recharge was calculated using different water balance and infiltration models, and for the two sampling stations. Soil and epikarst are modelled as a first compartment where evapotranspiration occurs, defined by its water capacity. This subsurface water capacity has a strong influence on the isotopic signature of the recharge. It may contribute to deplete the isotopic value of regional groundwater. Moreover, time variation of groundwater isotopic signature provides information on the functioning of the aquifer. Finally, water stable isotopic signatures of local groundwater are identified and plotted in a synthetic graph useful for karst groundwater management.
Transit time can be estimated thanks to natural tracers but few of them are usable in the 0-6 months range. Blondel (2008) established a relationship between HIX (humification index) and transit time of less than 1 year. HIX corresponds to the ratio of organic matter with a heavy weight (less metabolized by microorganisms) to light weight (highly metabolized; Ohno, 2002; Zsolnay, 1999). The goal of this work is to analyse and improve the use of this type of ratio as a natural tracer of transit time.
In this purpose, we realised a critical analysis of previous works. Careful review of the work of Blondel (2008) shows that this analysis suffered of methodological bias. HIX was first developed by Zsolnay (1999) using 2D spectra. Ohno (2002) showed that Zsolnay’s HIX needed to be corrected for concentration influence. He thus proposed a modified formula that directly accounted for concentration influence. Blondel (2008) built the relationship between transit time with the initial equation from Zsolnay (1999), thus not accounting for concentration influence, and using a different emission wavelength than that of the reference paper. The relationship used to link transit time and HIX by Blondel (2008) was also based on a very limited number of samples.
A more fundamental remark is that HIX has been first developed for soils and that the wavelengths used to characterize heavy and light weight organic matter is not adapted to groundwater dissolved organic matter types.
It thus appears that although the ratio of heavy and light organic matter does seem to be a good candidate to be used as a natural tracer of transit time, this approach needs to be rebuilt.
We conducted a one-year campaign (june 2020 to june 2021) of bi-monthly measurements of major elements, TOC, stable isotopes, excitation-emission matrix (EEM) and 2D spectra of natural organic matter fluorescence on two karst springs from the Fontaine de Vaucluse system (Saint Trinit and Millet springs). Both are the outlets of small karst systems with a high karstification level. Saint Trinit reacts faster and has shorter mean transit time than Millet (Batiot et al., 2003).
We first identified excitation-emission wavelengths windows of humic-like organic matter (heavy weight) and protein-like organic matter (light weight) based on our set of EEM. We then defined the Transit Time index (TTi) as the ratio of humic-like organic matter to the sum of all fluorescent compounds (protein-like and humic-like).
Mean value of TTi indicates a longer transit time for Millet than St Trinit, in agreement with the hydrodynamic behaviour of those springs. HIX as defined by Ohno, 2002 yields opposite results.
These promising results hint that TTi is in the process of becoming a natural tracer of transit time for karst springs.
Batiot, C., Emblanch, C., Blavoux, B., 2003. Carbone organique total (COT) et magnésium (Mg2+) : deux traceurs complémentaires du temps de séjour dans l’aquifère karstique. Comptes Rendus Geosci. 335, 205–214. https://doi.org/10.1016/S1631-0713(03)00027-0
Blondel, T., 2008. Expérimentation et application sur les sites du Laboratoire Souterrain à Bas Bruit (LSBB) de Rustrel – Pays d’Apt et de Fontaine de Vaucluse. Thèse de l'Université d’Avignon et des Pays de Vaucluse.
Ohno, T., 2002. Fluorescence Inner-Filtering Correction for Determining the Humification Index of Dissolved Organic Matter. Environ. Sci. Technol. 36, 742–746. https://doi.org/10.1021/es0155276
Zsolnay, A., Baigar, E., Jimenez, M., Steinweg, B., Saccomandi, F., 1999. Differentiating with fluorescence spectroscopy the sources of dissolved organic matter in soils subjected to drying. Chemosphere 38, 45–50. https://doi.org/10.1016/S0045-6535(98)00166-0
The Cadarache Center for Atomic Energy and Alternative Energy (CEA) is one of the largest energy research and development centers on a European scale. Located in the French Mediterranean region, the site is subject to significant rainfall events such as that of November 2011 with a cumulative rainfall of 225 mm in 72 hours. These intense events can lead to groundwater flood with rapid kinetics and high amplitude piezometric rises (several tens of meters in a few hours). In November 2011, the increasing of the Cretaceous limestone aquifer level reached 40 meters, depending on the location. The 1,600 hectares center has an outstanding monitoring network including 8 rain gauges and more than 400 piezometers, of which around 230 are equipped with pressure sensors.
In order to improve the groundwater flood forecast, it is necessary to have a better understanding of the hydrodynamic behavior of the Cretaceous aquifer. The goal of this study is thus to improve the understanding of the aquifer’s piezometric responses, and its heterogeneity. The method used was the following.
The results are the following: four clusters were identified; three of them appeared to be representative of specific areas. For each specific area, piezometers were overlaid on different plurikilometric NW-SE and NE-SW lineaments identified by previous bibliographic studies.
This extensive study on nearly 80 piezometers (for which data are available during the studied events) has thus allowed to define a method of automatic feature extraction of the sorted groundwater levels, and to establish a clustering of piezometers according to their statistical behavior. Even if the statistical "black box" method was applied without any prerequisite on the location of the boreholes, it allowed establishing a very relevant relation between sorted water height and geological structure of the flows. This preliminary study will be complemented by considering other factors such as the type of limestone in the different zones, or information on the non-satured zone. In addition, other statistical analyzes are envisaged on the piezometric data of the center, possibly on other data as temperature or electrical conductivity.
Regarding the flood hazard, it will be interesting to complete the statistical analysis with long term pumping tests.
Different modeling approaches from black- and grey-box to distributed methods can be employed for forecasting observed state-variables of karst systems. Considering this objective, models can be straightforwardly compared via statistical fitting metrics, such as Nash-Sutcliff and Kling-Gupta efficiencies. However, every model type comes with some capabilities and limitations that may or may not fulfill other modeling objectives, e.g., system characterization. This work focusses on multiple linear and nonlinear regression (MLR and MNLR) and transfer function noise (TFN) models as two time series analysis tools for karst spring discharge simulation. MLR model is based on the identification of relevant lags from the autocorrelations. A stepwise approach was here used to calibrate spring discharge (as response) based on the identified lags of discharge and climatological time series (as predictors). The MNLR model additionally explores relevant second-order predictors and interactions. TFN modeling utilizes pre-defined system response functions to represent the transformation from system input to system output. We utilize a two-flow-component nonlinear recharge model coupled to the TFN model to simulate spring discharge.
The Milandre karst system, a well-studied and -modeled system in Switzerland was considered as the case study, where our approaches can be compared with the distributed discrete-continuum model MODFLOW-CFPv2 and 12 other models of different types, conducted within the framework of Karst Modeling Challenge (KMC). In KMC, similar data-sets of climatic variables (i.e., hourly precipitation, and temperature, as well as daily potential evapotranspiration at one to three climatological stations) and hourly spring discharge for two periods in 1992-1995 and 2014-2015 were provided for calibration to forecast spring discharge for 2016.
In general, all employed time series models showed a good fit to the calibration dataset, while it was tried to avoid model overfitting via the usage of relevant statistical measures. For example, the prediction r-squared (pred. R2) for all model calibration periods was close to unity for MLR and MNLR models, showing how well the models may forecast responses for test periods. The TFN model approach has an additional benefit of conceptually reflecting some of the real system processes while keeping the number of parameters moderate to low.
Results show that the MLR, MNLR, and TFN models outperform 4, 4, and 13 out of 13 KMC models at forecasting, considering the fitting metrics, respectively. In general, MLR and MNLR are simpler than TFN and many other models employed in KMC. Although they can prove the relative importance of former events and system memory on the observed behavior, they do not provide much system understanding. The linear approximation of the karst system internal flow processes from recharge to discharge with transfer functions has been shown to produce good fit metrics for calibration and forecasting. However, the system nonlinearity gets in turn represented as part of the recharge process from which the physical interpretability of the model suffered.
Although the employed data-driven models statistically outperformed the process-based CFPv2 model with lower computational effort, they could not help characterizing the flow system, e.g., the conduit network.
About 40% the world’s population depends on rivers originating in high mountains for their water supply. Often, groundwater contributes substantially to mountain streamflow and represent an important water source. Therefore, it is important to understand the groundwater storage and discharge processes in mountainous environment. However, such systems are usually characterized by great geological complexity and highly heterogenous hydraulic properties. For that reason, proper system characterization, monitoring and modeling remain challenging. In this study, we investigated a geologically complex alpine catchment in the Dolomites (Italian Alps) by combining hydrogeological mapping, hydrological monitoring and numerical modelling. A lumped parameter model by combining linear and non-linear reservoirs was applied to simulate the continuous measured catchment discharge in a period of three years, which covers a large variation of hydrodynamic conditions. The current model structure can consider the main surface and subsurface water storage and discharge processes within the studied catchment: (1) snow accumulation and melting, (2) karstic conduit and matrix flow in a fractured dolomitic aquifer and 3) flow through the unconsolidated deposits (porous aquifer) accumulating on the slopes and at the valley floor. In order to evaluate the model realism, we conducted a Regional Sensitivity Analysis for main model parameters. The current results demonstrate that the newly developed model can reproduce most hydrogeological variability of the catchment. Besides the matrix and conduit flow in fractured dolomitic aquifer, they also highlight the important role of unconsolidated sediments (porous aquifer) to the storage and discharge behavior of the entire groundwater system.
Young researchers meeting
Simulation of flow physics in fractured-karstified carbonates is often hampered by the difficult discrete representation of the multiscale heterogeneity of these media, whose representative elemental volume often exceeds that conventionally studied in the laboratory. Under these conditions, how and at what scale can the petrophysics of the carbonate matrix, fractures and/or karst be quantified? What rules of homogenization should be applied to upscale from the measurement of the media towards their representation in the model? This problematics reaches its paroxysm in unsaturated zones for which a relative permeability and capillary effects should be considered, but is only rarely addressed in the literature.
This study is part of a wider research project dedicated to the understanding of fluid dynamics at various scales of a karstified carbonate reservoir with respect to its multiphase geological history. Here, we are interested in the characterization of the petrophysical properties of carbonate facies from centimetric to metric scale.
Several boreholes a few meters apart were cored and exhaustively characterized. Petrophysical measurements were performed on rock samples taken from the cores while water injection tests between packers were used to investigate some metric intervals of the unsaturated medium surrounding the wells. In this study, these hydraulic tests are interpreted using both analytical solutions and numerical models.
The sensitivity of the results to the analytical solutions used in the interpretations, to the actual experimental conditions, and to the simplified representation of the medium is investigated through numerical experiments. In particular, the representation of each rock interval tested by a single or double porosity medium is analyzed. The petrophysical parameters obtained by inversion of the tests are compared to the measurements made on the rock samples. The results from the analytical approach are also compared to those from the numerical approach, with respect to the common or specific assumptions of each method. The numerical model also allows to visualize the probable spatial and temporal evolution of pressure and saturation during the test to better understand the processes involved.
In the absence of an analytical solution dedicated to the interpretation of tests in unsaturated zones, saturated zone solutions have been successfully applied and sensitivity to the initial saturation of the medium has been numerically evaluated. For most of the tests, the interpretation requires a dual-medium solution to properly fit the analytical solution to the actual data. In most cases, the matrix properties obtained are consistent with the measurements performed on centimetric samples. The permeability contrast between the two media can be related to the characteristics (nature, density, opening) of the discontinuities or the high frequency alternation of different rock-types observed on the tested intervals. Even if the protocol used investigates small volumes and makes it difficult to parameterize the water retention model, coupling analytical and numerical approaches opens insights into a new test protocol in unsaturated media.
Quantifying karst water stock in vadose zone is highly complex due to heterogeneity and the limited spatial representativity. Evapotranspiration (ET) is a major term in the water cycle which accounts for 70% of the continental precipitation returned to the atmosphere. However, it remains one of the biggest sources of uncertainty to balance water budget and particularly in karst area. It is necessary to better understand this component and to develop new approaches to improve its quantification. We explore how gravimetry can help to link ET to water stock in karst vadose zone. Gravimetry is a geophysical method that measures variations in the Earth's gravity field “g”. Superconducting gravimeters (SG) provide continuous and precise measurement of g which can be related to changes in the water storage with an accuracy of few millimeters. We seek to evaluate the potential of this approach to estimate the ET from daily time series within a large karst system. The Low Noise Underground Laboratory (LSBB) has been chosen because it is the only site in the world to have two vertically distributed SG: one on the topographic surface and one near 500 m deep. This unique configuration allows to subtract the deeper signal to the upper one in order to obtain a surface to depth residual where all global effect (e.g., tides, polar motion, atmospheric loading) were removed. This residual signal allows us to observe most of the seasonal variations in the water stock, which is located between the two SGs in the karst unsaturated zone. This very clean signal was then compared to ET values calculated from the SimpKcET model. In favorable conditions (i.e., free rain days, no gravity signal disturbance), we were able to observe a significant correlation between the daily changes in gravimetry and the outlets of the karst vadose zone (i.e., modeled ET and measured discharge). The daily variation in gravity signal minus discharge can be mainly attributed to the daily variation in ET. We therefore propose a way to indirectly assess ET by monitoring the vadose zone water stock with a large integration volume. However, this very promising approach is only valid under few conditions that are rain-free days with favorable weather conditions (e.g., no big pressure variations, no thunderstorms). This approach being valid with surface to depth residuals, it is therefore more difficult to observe this type of relationship with a single surface-based SG. This would require improving data processing in order to reduce noise. Determining ET from hydrogeodesy is a real challenge that approaches the technological and signal processing limits of gravimetry. Our ambition is now to work on single SG in order to reproduce and adapt our methodology on other experimental sites equipped with SG.
The classifications of the hydrogeological functioning of karst systems used today are mainly based on indicators derived from spring hydrograph analysis. From these hydrodynamic approaches, it is sometimes difficult to make a link between the hydrogeological response and the properties or structures of the aquifer. To overcome this issue a promising way is to use the additional signal of electrical conductivity (EC), which provides information on mass transfer and on the complexity of the processes involved (transfer velocity, contributions from unsaturated and saturated zones, etc.). The objective of this work is to propose new indicators in order to improve the classification of karst aquifers based on a combined analysis of rainfall-discharge-electrical conductivity (P-Q-EC) time series.
For this purpose, continuous hourly datasets of P, Q and EC over several years were collected on 21 French karst springs (SNO karst network and BRGM). A typology of these springs was carried out according to their hydro-climatic context, their structure (percentage of allochthonous recharge, importance of the saturated zone, etc.). In a first step, a statistical and signal processing analysis (frequency, correlation, and wavelet analysis) was used in univariate and bivariate approaches in order to produce different indicators giving insights on the hydrodynamic and hydrochemical responses of the aquifers. In a second step, a statistical approach was used to identify the indicators that differentiate karst aquifers according to their type. It was found that Q signal helps differentiate between karsts with a small saturated zone (Jurasian type) and karsts with a deep one (vauclusian type). On the other hand, the EC signal allows differentiating karsts characterized by a recharge from karst outcrops only (unary karst) from karsts influenced by an allochthonous recharge (binary karst). In a last step, a classification based on indicators derived from Q (memory effect) and CE (coefficient of variation) is proposed to differentiate karst aquifers according to their functioning and recharge type.
Finally, this work shows the added value of the information contained in the EC signal combined with Q to better differentiate the hydrogeological response of karsts in relation to their properties and structures. This gives insights for classifications of karst aquifers accounting for vulnerability in addition to the groundwater availability estimated from Q signal.
Amal MHIMDIª, Ines EZZINEª, Fadoua HAMZAOUIᵇ, Mohamed GHANMIª, Rachida BOUHLILAᶜ
ª Laboratory of Geoscience, Resource mineral, Energy and Environmental (LGREE), Geology Department, Faculty of Sciences of Tunis, University of Tunis El Manar, Tunis 1060, Tunisia.
E-mail : mhimdi.amal.c2i@gmail.com
ᵇ Laboratory of Sedimentary Basins and Petroleum Geology (SBPG), LR18 ES07, Geology Department, Faculty of Sciences of Tunis, University of Tunis El Manar, Tunis 1060, Tunisia.
E-mail : fadoua_fst@yahoo.fr
ᶜ Department of Civil Engineering, Modelling in Hydraulic and Environment, Laboratory, National Engineers School of Tunis, University of Tunis El Manar, Tunis, Tunisia.
E-mail : rachida.bouhlila@enit.utm.tn
Abstract:
Karst aquifers belong to the fractured aquifer family. The Zaghouan region located in NE of Tunisia (North Africa) is characterized by a high degree of karstification due to the climate impact and the development of fracture network. Survey using electrical resistivity tomography (ERT) is deployed to provide a cost-effective characterization of the subsurface karst environments. A total of four ERT profiles with a length of 300 meters in length were evaluated in the governorate of Zaghouan.
The area represents an anticline of Jurassic limestone rocks, which is overlain by a thin clay layer. In this study, an ERT survey was conducted to examine the spatial distribution and shape of underground cavities in the karst area of Jebel Zaghouan and Jebel Ben Saidan. In this study, geological, hydrogeological and electrical resistivity tomography (ERT) methods were applied to determine the geometry of the karst aquifer in the Zaghouan area (NE Tunisia). The area is characterized by fractured and karstic limestone aquifer of Jurassic. Four resistivity profiles were carried out along the study area at two sites (Jebel Zaghouan and Jebel Ben Saidan). The orientation, extension and the degree of inclination of those profiles are shown in a location map. The correct resistivity data was interpreted using Earth Imager 2D software. The results of the interpreted geo-electrical sections showed that the resistivity of the carbonate aquifer ranges from 2.5 to over 5794 Ωm. The thickness of the aquifer varies between 20 and 50 meters, while its depth from the surface is between 10 and 60 meters. The ERT not only provided accurate near-surface information, but was also very useful in establishing the geometry of the aquifer. It was also very useful in establishing the 3D geometry and position of several potential cavities and karts. The results show the presence of large isolated cavities at different depths. The ERT technique could be used for 3D detection of underground limestone cavities.
Keywords: Zaghouan (NE Tunisia), karst aquifers, electrical resistivity tomography (ERT), cavities.
Several tools exist to simulate plausible karst network geometries but there is little information about the statistical distribution of the diameters of those conduits. Some networks have been mapped by speleologists and the distribution of the explored conduits can be analyzed, but the data usually do not cover the smaller innaccessible conduits, therefore the data sets are incomplete.
An alternative to overcome this lack of data is to run numerical speleogenesis model and study the statistical distribution of the karst conduits and its evolution during the life of a karstic system. In this presentation, we will introduce KSP (the Karstification Simulation Plug-in) a Feflow 7 plugin that was developped to simulate speleogenesis driven by the dissolution of limestone rock. The plugin allows to model the transition from laminar to turbulent flow, as well as the opening of the fractures in reaction to chemical reaction between the matrix and the groundwater. In a first step, the results of the plugin are validated against standard speleogenesis benchmarks.
In a second step, we consider 2 different vertical cross sections with recharge on the top and an outlet on the side of the section. The medium is made of an homogeneous matrix with different cases of discrete fracture networks with apertures simulated following a truncated log-normal distribution. The evolution of the karst conduit network was simulated under different boundary conditions (i.e. different initial hydraulic heads) and simulations were run beyond the breakthrough of fractures connecting inlets and oulets. The post-treatment of results was focused on assesing the evolution of of the karst conduit size distribution after initial breakthroughs.
The results show that the discharge rate at the spring often displays different phases of breakthrough corresponding to the evolution of the network. We observe the extension of a cave gestation zone that progress backward from the spring area. When the karst conduits in agestation zone fully developped, a steep increase in discharge rate is observed. As a consequence, the statistical distribution of the conduit aperture contains multiple modes corresponding to these successively propagation of the zone . When a zone is well connected to the spring, the main enlarged conduit in the zone is still continuing to grow while the surrounding conduits keep their original aperture. This is how the statistical distribution of the fracture apertures or conduit sizes becomes multimodal.
We argue that even if those results have been derived for some very specific conditions they can be used to select some reasonable statistical distributions of conduit sizes from the centimer to the decameter scale.
The groundwater represents a major drinking water reserve, at world scale. Among aquifers, the karst supplies about 25% of world population. The karst-type hydrosystem is famous for its high vulnerability, directly linked to its spatial heterogeneity and temporal variability. Those features, combined with anthropogenic stresses and issues related to climate change, impact the water quality. The full understanding of karst hydrosystem is needed for a better use of water resource, but it still needs large improvements.
TRANSKARST (TRANSdisciplinary research on KARST) is a research consortium, built in 2019 and gathering scientists and water resource managers around the study of the Arcier karstic spring (Eastern France). This spring supplies about 50% of the drinking water of Besançon city (117 000 inhabitants). By linking multidisciplinary studies, this project aims at better understanding the impact, the origin and the pathway of mineral, organic and microbiological contaminations affecting regional karstic aquifers.
Since February 2020, numerous data have been collected: structural geology, geophysics (ERT, gravimetry and passive seismics), hydrodynamics (rainfall, water level/flow rate), physico-chemical (electrical conductivity, temperature, turbidity, major and traces ions, isotopes) data and artificial tracer tests. The first part of the project concerns the hydrogeological and hydrodynamical modelling of the Arcier hydrosystem. With the combination of structural (Geomodeller 3D, 3D Visual Karsys) and functional (time series analysis, Kartsmod) modelling, we propose a comprehensive model of the hydrodynamical behavior. Then, the understanding of the hydrosystem functioning can be involve by hydrochemistry data.
The geological model of the watershed shows a complex structure with heterogeneous but spatially organized faults and fractures. Geological and geometrical offset along fault impact the permeability of the hydrosystem. Hydrodynamical model based on rainfall/flow rate analysis and tracer tests show a direct link between the geology and the karst reactivity, with a distinction between west and east catchment sides. Hydrochemical datas confirm this bimodal functioning. In low water period, chemical content of Arcier’s water is influenced by a “geological pole, determining by the eastern part of the water catchment, and an “anthropogenic pole”, mainly corresponding to the western part. In high water period, the distinction between the two poles disappears. In the long run, the model will be improved to predict the contamination circulations into the Arcier’s hydrosystem.
The Vise spring is the main outlet of a Jurassic karst aquifer located close to Montpellier city, South of France. The Vise spring is submarine, occurring at the bottom of the Thau lagoon at a depht of 30 m. The lagoon, made up of brachish water is connected to the Mediterranean sea. The fresh water from the karst aquifer as a whole, and especially from the spring, contributes to the qualitative state of the Thau lagoon, which is well known for shellfish farming activities.
During the last fifty years (from 1967 to 2014), six occasional saltwater intrusions (called “inversac” in French) occurred, inversing the water flow at the submarine spring during a period varying from a few weeks to a few months. This backflooding process at the spring induces a very large saltwater intrusion into the karst aquifer. Given that this aquifer provides several highly important ecosystem services (drinking water supply for the coastal villages, fresh water to the Lagoon, thermal water to Balaruc spa and health resort), a large program of groundwater monitoring has been recently launched.
An observatory has been installed over the territory since 2019. Offshore in the Thau lagoon, the submarine Vise spring was equipped with flow recording devices as well as electrical conductivity and temperature monitoring sensors. Onshore but close to the spring, three boreholes of 45 m, 168 m and 300 m deep each were drilled near an existing thermal borehole, and a new borehole including fiber‐optic distributed temperature sensing (FO‐DTS) is currently under completion.
Sub‐hourly observations of pressure, electrical conductivity and temperature in these boreholes and in the lagoon, as well as synoptic measurement campaigns focused on groundwater chemistry, complete a monitoring network of about twenty boreholes and springs spread accross the territory.
In november 2021, a seventh backflooding event started and is observed with the new monitoring system. From an initial flow rate of about 60 l/s from the aquifer to the lagoon through the spring, the flow inverted to about 350 l/s from the lagoon to the aquifer in a few minutes on the 28th November 2020 at 9:40 AM. This sudden backflooding created a sudden water level rise of about 2.5 meters into the karst confined aquifer. A few months later, the saltwater intrusion into the aquifer is still high, equal to 150 l/s.
A physical mechanism is proposed to explain the sudden inversion of flow and its long duration after that the event has started. The propagation of a piezometric wave through the aquifer is simulated using simple analytical solutions. A preliminary meshed model of the aquifer and its interactions with the lagoon has been developed for identifying the main processes. First attempts for proposing alert indicators are also discussed.
Spatial characterization of the hydraulic properties in the subsurface is an extensively studied problematic. Inverse problems allow to image those properties by interpreting the information from a dataset of field measurements with a chosen physical formulation of fluxes in a numerical distributed model. However, karst media characterization remains a complex task, due to the fact that the matrix and conduits entities generate a highly contrasted distribution of property values. Furthermore the matrix and conduits compartments respond to different flow physics. Thus, one needs to employ a multi-physics model, an inversion method able to represent the properties contrast and also use data providing information on the localization of the conduits network and its connectivity.
We propose a large-scale 2-D application of characterization of the Lez aquifer in southern France, covering a surface of about 400 km². We take advantages of long-terms measurements within the framework of the MEDYCYSS observation site, part of the Karst observatory network (www.snokarst.org) initiated by the French institute INSU/CNRS. Drawdown signals measured in 11 wells and incorporating a periodic response due to a daily pumping at the aquifers spring were thus considered to identify the location of the conduit network. The periodic responses can provide connectivity information between wells in the inversion process, while non-periodic responses will permit to better assess the large-scale property values of the whole hydrosystem. A Cellular Automata-based Deterministic Inversion (CADI) is used to generate a contrasted property field able to reproduce the measured signals in the 2-D distributed numerical model solving Darcy flows. However, pressure data alone remain limited to characterize the fast flows that can occur in the conduits network. Thus, the flow velocities in the preferential flow paths located with the pressure data are then reconstituted by inverting a set of different tracer tests responses at the Lez spring, considering this time a pipe flow physics in the model.
Abstract
The hydrodynamic response of carbonate aquifers is conditioned by the distribution and development of the karst conduits network within the aquifer, as well as by the intensity and relevance of the recharge processes. In the Ubrique karst system (southern Spain), recharge takes mainly place by two mechanisms (binary karst system): by direct infiltration of rainwater over the bare carbonate outcrops (autogenic component) and by means concentrated infiltration of runoff water coming from small neighbor catchments formed by low-permeability materials (allogenic component). Groundwater discharge occurs along the southwestern border of the carbonate outcrops, nearly the Ubrique urban area, through two perennial springs (SP-1 and SP-2) and several overflow springs, although there is only continuous record of one of them (OFS-2).
In this work, mean response times obtained by cross-correlation analysis (hourly series from 2013 to 2018) have been compared with the experimental results derived from a dye tracer test carried out in March, 2018, in order to advance in the understanding of the system behavior, particularly to assess response times to several scales. The first methodology gives information on the global –mean- response of karst systems concerning recharge events, analyzing the relationship among the input (rainfall) and the output (spring discharge) time series. On the other hand, dye tests provide quantitative information on the aquifer behavior (mean flow velocity and transit times) under specific recharge conditions and during particular flow conditions.
Results from the long-term analysis showed a mean response time-lag of 58.2 hours for SP-1 spring and values of 68.6 and 81.4 hours for OFS-2 and SP-3 springs, respectively, taking as the input signal reference the same rainfall station. On the other hand, the shortest dye detection times in the outflow points were 38.5, 55.2, and 64.3 hours for SP-1, OFS-2, and SP-3, respectively. Degree of karst development, hierarchy in the conduit network, or the presence of low permeability outcrops imbricated along preferential drainage paths, provoke characteristic response times in each spring, which affect both to the responses to a particular recharge event and to the system functioning as a whole. The strong relationship of the different methods applied can be used to validate the representability and accuracy of statistical approximations, to establish a plausible conceptual model for the aquifer behavior, and to the implementation of protection zones and the validation of groundwater vulnerability maps. Nevertheless, for a robust interpretation of the hydrogeological functioning of karst aquifers is necessary to consider several and complementary hydrogeological research techniques.
The freshwater supply and protection constitute a major challenge in the context of global change. Among other freshwater stocks, karst aquifers constitute heterogeneous hydro-systems with non-linear hydrogeological behavior. The high level of heterogeneity in such aquifers and the lack of knowledge of their internal structure lead to strong difficulties in the management of their freshwater resource. Facing such difficulties, karst aquifer modeling meets difficuties [1] to design suitable decision-making tools concerning both quantity and quality aspects of freshwater sustainability and [2] for flood prevention and flood forecasting purpose. Then, coupling various modeling approach provides a deeper understanding in karst hydrology as well as more reliable numerical modeling. We propose a synthetic description of the various modeling approach currently developed at the HydroSciences Montpellier (HSM) laboratory for the flow and transport simulation across the different spatio-temporal scales. Methods are developed to account with processes on spatial scales from centimetric to kilometric and temporale scales from seconds to decades.
Concerning water quantity, various hydrological modeling approaches are applied can be applied for groundwater resource variability and spring discharge forecasting in a context of global change. Deep numerical investigations are needed to account with the uncertainties related to [1] hydrological models (i.e. structure, parameters), [2] fluxes quantification (i.e. rainfalls, evapotranspirations, discharge) and their spatio-tempral heterogenities or even [3] changes in anthropogenic pressures such as pumpings and land-use, related with a growing population.
Concerning water quality, various approaches are developed to simulate non-reactive solute transport with potential application to reactive transport and potential extension for pollutogramm simulation and Early-Warning System (EWS). This approach consists mainly in the developpement of innovative method for the interpretation of artificial tracer tests. In addition, natural tracers as wall as electrical conductivity or dissolved gazs provides insight about flow processes within the karst aquifers, and constitutes relevant data for hydrological modelling. Finally, coupling flow and transport simulation is assumed to give robustess in karst aquifers conceptualization and so in karst groundwater vulnerability assessment.
Karst Aquifers contribute to freshwater supply in several Mediterranean countries. Recharge zones of karst aquifers, which are often hydraulically connected over large areas, are highly vulnerable to contamination because of their hydraulic properties, such as rapid and turbulent flow in a network of conduits, resulting in highly variable spring discharge and water quality (Hartmann et al. 2014).
In Tunisia, little work has been done so far related to the study of karst using remote sensing and GIS technologies. Thus, this study aims to contribute to an inventory of karst features and to the detection of causal and triggering factors influencing their development and karstification processes, mainly based on remote sensing and GIS data in Djebel Zaghouan site, as much as possible. Located in North East Tunisia, the Djebel Zaghouan karst aquifer (area =19 km2) was naturally discharged by several karst springs and galleries. Nowadays, the aquifer is totally exploited by boreholes that supply drinking water for the local cities.
Therefore, in the present work we focused on identifying recharge zones and their evolution using remote sensing. The adopted method was tested in Morrocco (Theilen-Willige et al., 2014) In addition we applied APLIS method to produce recharge maps for the aquifer.
First, we made digital processing of a satellite data images (SENTINEL image level 2A date 25/09/2019 ) having multiple spectral band, with spatial resolution between 10 and 20m. Vegetation indices NDVI and NDWI were computed to identify vegetation sites presenting highest humidity. The DEM ASTER GDEM (30m resolution) was used to determine sinks depth zones as the difference between the raw DEM and the corresponding DEM with filled Sinks, a procedure that fills all the depressions of the DEM. The combination of sinks depth, slope, NDVI, NDWI layers and the geological map gereates the final potential sinks recharge map. The obtained sinkmap was processed to eliminate falsely identified "Doline" karst depressions using morphometric attributes.Comparison of the obtained map with those produced using APPLIS method and their validation with the observed dolines in the field is in progress
The patterns and surface alignments of karst features often are associated with joint patterns, faulting and folding. Conduits in karst groundwater are formed from rock dissolution along planes or discontinuities. Therefore, investigations of the relationship between the occurrence of dolines, karst development and the tectonic pattern are among the research issues.
Theilen-Willige, B.; Malek, H.A.; Charif, A.; El Bchari, F.; Chaïbi, M. Remote Sensing and GIS Contribution to the Investigation of Karst Landscapes in NW-Morocco. Geosciences 2014, 4, 50-72. https://doi.org/10.3390/geosciences4020050
Until the 80s tracer tests with fluorescent dyes were made with mechanical samplers. The sampling period extended typically between one hour and one day. Since then, these bulky devices have been advantageously replaced by the flow-through fluorometer, with sampling period as low as two seconds, and virtually no time limit. The concentration of three different tracers and the turbidity can be measured, recorded on flash memory and displayed in real time. However the method is limited by the selection of the tracers: For example, rhodamine-based tracers (amido-G, sulfo-B, WT) and eosine cannot be easily separated from a mixture, and it’s still worse if the turbidity exceeds 10 NTU. Another disadvantage: Poor separation of excitation light from fluorescence, due to limited optical filtration capability, causing the detector to glare. We present a new design of the flow-through field fluorometer in which a micro-spectrometer replaces the detection optics. The new method addresses these two limitations.
The optical spectrum of the micro-spectrometer extends from 340 to 850 nm, with a resolution of 8 nm. A laser diode produces the excitation of the fluorescence. The emission spectra of any dye tracers appear clearly separated from any other tracer fluorescence, and above all, from the excitation spectrum. A full spectrum is obtained within seconds. For long-lasting tracer tests with sampling period in the minute range, the detection level can be lowered in proportion to the integration time of the CMOS sensor.Examples with various tracers are shown.
To develop an adapted and resilient strategy to characterize the exchanges between the Ardèche River and three associated groundwater bodies, a research project has started in 2021. The surface of the watershed area is roughly 2,400 km2. These water bodies (Triassic, Jurassic and Urgonian) are located in the southern extremity of the Gard department and develop essentially in the Ardèche department, on the edge of the Cevennes massifs in France.
In a previous studies on the Cèze River (Eurokarst 2018) two representative poles of the stream and aquifer have been determined and combined with principal component analysis (PCA). For this new project we use only PCA to characterize the chemical facies of the waters sampled on site in order to determine which ones are representative of each groundwater body.
This methodology allows to visualize the influence of groundwater bodies on surface waters by highlighting the chemical evolution in stream from upstream to downstream as it flows through the different water bodies.
The results of this first approach should allow to determine the contribution rate of groundwater bodies to surface water and the evolution of this ratio according to the seasons.
One of the missions of Andra, the French national agency for radioactive waste management, is to study and develop solutions for the sustainable management of radioactive waste that does not yet have a final disposal method. In this context, Andra is conducting research to design a deep geological repository for long-lived high and intermediate level radioactive waste. This project, called Cigéo (Centre Industriel de stockage GÉOlogique), should be built in the Meuse - Haute Marne interdepartmental area near Andra's underground research laboratory near the town of Bure. As part of the impact study of the surface infrastructure facilities associated with the Cigéo project, Andra is seeking to refine the characterization of the flow dynamics within the Barrois limestone, a Tithonian-age alternation of limestone and marls that is visible in outcrops. Overlying the Kimmeridgian marls, the Barrois formation is composed of 5 main lithological units: Sublithographic limestone, Pierre Chaline, Dommartin limestone, Oolithe de Bure and Carrié limestone.
Previous studies made it possible to characterize the lithological properties of this alternation of more or less karstified limestone and marl levels, but the complexity of the Barrois limestone implies to improve the characterization of its hydrodynamic properties and hydrogeological functioning (directions of underground flows, interactions between surface water and groundwater) at the scale of Cigéo as well as at regional scale.
The objective of this study is to characterize and understand the circulations within the Barrois limestones hydrosystem, which constitutes the most superficial geological formation at Cigéo scale and will thus be directly impacted by surface installations (buildings, chutes, access shafts).
For this purpose, a follow-up of hydrochemical measurements coupled with hydrogeophysical measurements has been set up on this site for 1 year. These non-invasive investigations complete and spatialize the geological knowledge acquired based by drilling.
Hydrochemical monitoring is performed at 3 sites. Electrical conductivity, temperature, flow rate, and natural fluorescence of organic matter are recorded at high frequency (< 1 day). Water samples are also taken every 4 days for major elements analyses. This hydrochemical monitoring aims to qualify and quantify:
- The signal of infiltration of water surface into the underground system by losses, bearing anthropic and lithological (marl) indicators at the head of watershed;
- The groundwater through the sublithographic limestones at the mid-course of the watershed,
- The groundwater outflow, integrating all the flow of the aquifer.
An initial hydrogeophysical investigation campaign was carried out during the 2021 summer (low water conditions) using magnetic resonance sounding (MRS). MRS measurement sites were selected based on ambient electromagnetic noise and lithological conditions. The aims of this hydrogeophysical campaign were to quantify water stock variations in the main aquifer of the sublithostratigraphic limestone and make hypothesis on the flow conditions (drainage or storage).
The cross interpretation of the results of the hydrochemical temporal monitoring and the hydrogeophysical campaign suggest that the tend to show that the functioning of the sublithostratigraphic limestone aquifer in the sector of head of watershed concerned by the Cigéo project is very reactive and not very capacitive.
The perspective of the work is to spatialize the geophysical measurements on two zones distributed on the downstream of the catchment area and to confront them with the hydrochemical monitoring. These measurements will be reproduced under different hydrological conditions.
Karst systems are complex hydrodynamic systems, characterized by slow flows in the matrix compartment and quick flows in the conduits. This leads to a very specific management of this resource in terms of quantitative and qualitative risks. Different approaches exist in order to identify the quick flows connectivity at large scale or the interactions between the slow and quick flows compartments. Among these approaches tracer tests remain easy-to-implement methods, consisting in injecting a tracer in the groundwater at a location and measuring its recovery at a spring downstream. This approach is used in a similar way to estimate surface flowrates. However, when applied to surface flows, one takes the morphology of the river into account, which is impossible in underground applications. Therefore one way wonder: what uncertainties arise from neglecting the conduits morphology when interpreting breakthrough curves for tracer tests in underground karst systems?
We provide elements of responses to this question by performing an underground tracer test experiment in a mapped karst conduit. The experiment is led at the Fontaine de Nîmes spring in collaboration with roboticist, electronician and diver colleagues in the framework of the ‘Lez2020’ ERDF research project. A robot is guided through the flooded conduits by divers and performs an acoustic 3D imagery of the conduits walls. It is then fixed in-situ approximately 250m upstream from the spring where it autonomously release a mass of tracer. The breakthrough curves of this test are then measured at different locations. A 3D numerical model is built to simulate the flows in the conduits, taking into account their precisely mapped morphology. The tracer tests is reproduced numerically and compared to the measurements in order to provide better insights to the effects of the conduits morphology and the turbulence it can induce at the walls on the tracer tests recovery responses.
Classification is a popular tool to gain insights into the characteristics of a karst system. Last year, we proposed a new classification of karst systems hydrological functioning based on recession curves analysis (Cinkus et al., 2021). The typology consists in six classes and is based on three aspects of a karst system functioning: (i) capacity of dynamic storage, (ii) draining dynamic of the capacitive function, and (iii) variability of the hydrological response. The methodology only requires karst spring discharge time series and can be applied with few years of monitoring.
We developed an interactive R application to support the proposal of classification and help with the completion of the analyses. In KarstID, the user can import and work on a karst spring hydrograph with multiple hydrodynamic analyses: statistical, recession curves, correlational and spectral, and classified discharges. The application heavily facilitates the application of the recession curves analysis, as the interface is interactive and let the user select the recession curves with a cursor. The other analyses are performed automatically when the data is imported, although the user can define the cutting point for the correlational and spectral analyses. The application also provides the classification of karst systems hydrological functioning based on the proposal of Cinkus et al. (2021) and offers to compare the results with a database of 78 karst systems located worldwide.
KarstID is developed in the R Shiny framework – which allows to build interactive web interface from R – and is embedded into an R package. It prevents the user to have any knowledge in programming except for the installation R packages. The KarstID package is open source, actively developed and available on Github (https://github.com/busemorose/KarstID). Further perspectives to this work are to (i) add more recession models, (ii) consider other hydrodynamic analyses, and (iii) discuss with the community to improve and refine the application.
In the Mediterranean area karst aquifers represent a primary water resource for human populations. This is due to the climate and the scarcity of perennial streams on carbonated watersheds. Many Mediterranean cities are predominantly supplied by karst water (Kresic and Stevanovic 2010). Nevertheless, the complexity and heterogeneity of these media hinder the characterization and modelling of flow processes. Karst springs have highly variable discharge and water quality, which induce a high vulnerability to pollution due to complex interactions between the hydrological and hydrogeological compartments. The understanding and prediction of water resource availability and water risks related to climatic and environmental changes in the Mediterranean context need to be studied at various time (flood event, season, inter annual variability) and spatial scales (local to regional) and thus require the development of specific approaches. To study this critical karst zone, an integrative approach based on coupled hydrological, hydrodynamical and hydrogeochemical tools is needed.
The Lez karst system is a relevant example of a typical Mediterranean karst aquifer. It represents a strategic water resource for Montpellier and its surroundings (southeastern France). Groundwater is abstracted directly from the main karst conduit upstream from the Lez spring, under active management since 1981. The Lez karst system is a part of the MEDYCYSS observatory (Multi scalE observatory of flooD dYnamiCs and hYdrodynamicS in karSt), of the French National Network of Karst observatories (SNO KARST, OSU OREME-IR OZCAR). Since 2005, regular and continuous monitorings of the hydrochemistry and hydrodynamics (wells, karst network, permanent and temporary springs), as well as of the river and spring discharges, have been carried out on this karst system. Meteorological variables and soil water content are also monitored at various locations. Monitoring of the water and CO2 exchanges between soil/vegetation/atmosphere have recently started.
The use of hydraulic tests at various temporal and spatial scales highlights the heterogeneity of the aquifer hydraulic properties. The hydraulic connectivity at both local and regional scales is mainly due to sub-horizontal flow-bearing structures, in which a conduit network has developed. This quantification of hydrodynamic parameters provides important constraints on multiscale modeling and characterization of the main flow paths in such a karst system.
The development of specific tools such as natural fluorescence, noble and anthropic gases by in situ continuous monitorings brings to light the complex mixing of fluxes flowing at the Lez spring and the aquifer’s vulnerability because of waste-water (raw or treated) infiltration through the temporary streams of the basin. Antibioresistance of bacterial communities has been also studied and stresses the importance of elucidating the ways in which hydrogeology and human practices impact the diversity and dynamics of microbial communities in karst groundwater and impact human health.
The various approaches and tools developed improve knowledge of Mediterranean karst systems and thus help estimate climate-change impacts on water resources. These results will give useful information to water suppliers and stakeholders in order to improve the management of freshwater resources.
References
Kresic, N., Stevanovic, Z. (2010). Groundwater Hydrology of Springs. Engineering, Theory, Management, and Sustainability. Elsevier, ButterworthHeinemann, Oxford.
Karst aquifers are well known as dynamic and heterogeneous hydrogeological systems. Underground drainage conduits formed by dissolution processes make these aquifers hard to be characterized. Thus, the incomplete knowledge of the conduits network obstacles the management of groundwater resources in karstic systems. Moreover, the uncertainty of the connection between conduits and porous media is the main problem reported by the scientific community and groundwater managers. In addition, these problems become more complex when the aquifer is subjected to intense exploitation. To design a sustainable exploitation plan, it is essential to characterize the aquifer behaviour, to quantify the available resources and to simulate the aquifer functioning under different scenarios (both climate and management). The assessment of these tasks is frequently supported by groundwater modelling tools. However, in karst aquifers numerical models are much more difficult to stablish than in detrital aquifers.
In this framework, research about how to simulate turbulent flow through real karst systems is still nowadays an open line research. One of the methods developed for the worldwide known code MODFLOW is the Conduits Flow Process (CPF) package, developed by United States Geological Survey - USGS. This package may simulate both turbulent and laminar groundwater flow that occurs in karst aquifers. In the present work CPF was applied to the Sierra de Becerrero carbonate aquifer (South of Spain), in order to evaluate its power as a useful tool for water management. The karstic conduits network discharges to three main springs (Ojo de Gilena, Ojo de Pedrera y Fuente de Santiago) in natural conditions. Exploitation wells are near the springs location. The Sierra de Becerrero model simulates the hydrodynamic behaviour (piezometric head, recharge, spring discharge, and storage changes during 30 years (1977-2006). CPF package allows to take into account karstic conduits network when they are active (steady-state or low exploitation conditions). Under overexploited or drought conditions, groundwater levels drops more than 20 m deactivating the flow along karstic network as a main flowpath, drying the springs and activating a preferential flow through the porous rock matrix. Three models have been developed, one in steady-state regime and two in transient state (including and not including groundwater extractions). Steady-state model reveals that CHD package must be applied to simulate the springs when the model run by CFP MODFLOW. Both flow path and available water resources rates seems to be altered under transient conditions when anthropic extractions are considered, showing the high influence of overexploitation. The transient model without well exploitation revealed changes in hydraulic parameters depending of the exploitation regime.
Based on this investigation of the hydrological behaviour of the Sierra de Becerrero aquifer for the studied period, it is recommended as future activities to update the model to the present and simulate possible future climate and management scenarios that could affect the system.
Sarkhun Spring emerges from the Ilam-Sarvak limestone formation on the Kerman Fault at an elevation of 1512 m asl in the center of the small town of Sarkhun, Zagros Mountains, Iran. The discharge of this spring was measured between April 2017 and June 2018. The maximum recorded discharge of 5730 L/s-1 occurred at the end of April 2017 and the minimum at the end of January 2018 was 410 L/s-1. A severe drought in late 2017 explains the lowest flow. The average discharge of the spring is probably ~1.0 m3 s-1. Substantial fluctuations in the spring discharge indicate the essential role of rapid (pulsed) groundwater flow. This spring is an important regional source of water for agricultural and domestic use. In recent years, it has suffered from significant oil pollution. The exact source of the contamination is not known, but the most important hypothesis is pipeline fracture along Gandomkar Fault, which is known to be active. Geological studies, hydrogeology and dye tracing were used to investigate the likely source of contamination. Uranine dye tracing, by injection into one of the fault branches that is six kms distant from the spring, yielded a velocity of about 500 meters/day along the trend of the Fault towards Sarkhun Spring. The great depth to groundwater at the fracture site (more than 200 m) and the presence of large conduits at the fault site make it difficult to remediate the aquifer. In particular, after the present study, there was a further fracture of the pipeline and a large volume of crude oil flowed into the fault path and out to the open surface (December 2020).
Karst aquifers are complex and heterogeneous. Given the importance of the water resources they contain, the sustainability of these aquifers requires the use of adequate tools and methods, especially for the mediterranean and semi-arid regions. The present study is part of the KARMA project and it focuses on Djebel Zaghouan, one of four studied sites around the Mediterranean sea.
Djebel Zaghouan limestone aquifer is one of the most important water resources of good
quality in the region. Currently, the aquifer is exploited by mainly 9 boreholes and galleries intended for the drinking water supply of the city of Zaghouan and the surrounding rural agglomerations. Three of these wells are used as commercialized mineral water.
The main objectives of this research work were to (1) improve the understanding of the hydrological behaviour of the karst system and (2) contribute to the development of a conceptual model (using the KarstMod software) that can be used later for management and climate change scenarios
Based on a previous modelling study [Nazoumou Y., 2002] of the karst aquifer of Djebel Zaghouan, we were able to collect information that led to the preliminary understanding of the site. Available Graphical discharge time series at the spring of the Nymphea dating from 1915 to 1927 at irregular time scales, were also collected and digitized.
The analysis of the available data showed that the transit time of the water flow was averagely long as the response time at the spring ranged between one to two months after a precipitation event. In fact, the karstic system of Djebel Zaghouan presented very inertial and steady hydrological functioning.
Several possible configurations of the model were tested on KarstMod, in order to consider the different possibilities of internal flows linking the reservoirs. For the same purpose, the ranges of the model's parameters were set to have a wide amplitude. We also considered different calibration and validation periods.
We observed that the simulations that satisfy the best performances have in common a basic structure where the three reservoirs E, M and C are activated.
Our model successfully reproduced the observed discharge fluxes with Nash–Sutcliffe efficiency coefficients (NSE) of about 0.83 and 0.72 for the calibration and the validation periods, respectively. However, few flood events were underestimated. In further work, the calibrated will be tested for time series corresponding to periods when the system was entirely exploited by boreholes.
Karst aquifers are characterized by intrinsic anisotropy resulting in a heterogeneous water flow system. In mountainous areas, the geological structure in orogens can also prompt hydrogeological complexity affecting groundwater flow distribution. However, to elucidate the influence of the tectonic factor in karst behavior is very often a challenging issue. In this work, several multitracer –dye- tests were performed in a tectonically complex alpine karst system located in S Spain. The pilot site hosts the major tectonic suture between the main geological domains of the Betic Cordillera (External and Internal zones). Aquifer recharge takes place from direct infiltration of rainwater through carbonate outcrops. Groundwater drainage occurs naturally from several permanent springs located at the southern border of carbonate outcrops belonging to the Internal Zone (defining the base level of the aquifer) and also by two overflow springs at the External Zone, in which water table is accessible in any hydrodynamic conditions (a vertical cave 30 m deep). Seven dye tracer tests were conducted in outcrops of the External Zone domain, between November 2015 and March 2018, involving different focused infiltration features (sinkhole, karrenfields and sinking streams; even an artificial one through a borehole) and flow conditions (high to intermediate). Results demonstrate the existence of fast flow karst connections (flow velocities up to 600 m/h) though the Jurassic carbonate rocks of the External Zone, defining a well-developed karst network, which is related to the bedding planes and first-order enlarged-fractures. Tracer tails detected into the shaft when springs was not overflowing suggest the existence of groundwater flows below the upwelling level, towards the saturated zone of the aquifer. However, injected tracers were not detected at the permanent discharge points located at the Internal Zone, which present delayed and buffered hydrodynamic responses to recharge events. We hypothesize with 1) a markedly karst development in the unsaturated zone of the Jurassic carbonate sequence; 2) a preferent participation of the unsaturated zone over that of the saturated zone during recharge conditions; and 3) a tectonic complexity in the contact area that interferes the flows of the saturated zone. From the methodological point of view, tracing techniques are very often limited when dyes are not detected at the discharge points, as occurred in the springs located in the Internal Zone. This could be consequence of a lesser degree of karstification into a sector of the aquifer, or to the existence of clayey materials of low permeability retaining the groundwater flow, even the tracers, or else because the saturated zone is of great thickness, or it has high capacity to homogenize. Nevertheless, for a robust interpretation of hydrogeological functioning in tectonically complex areas is necessary to consider diverse and complementary hydrogeological techniques (multi-criteria approach).
In karst groundwater, colloidal and suspended particles are ubiquitous and important vectors for metal transport within aquifers due to sorption processes onto colloidal surfaces because of electrostatic forces. Thus, the understanding of metal and sediment transport mechanisms constitute an important approach for characterizing groundwater origin and tracking principal flowpaths in karst systems where allogenic inputs substantially impact on groundwater discharge.
Sierra de Ubrique carbonate karst aquifer (Cádiz province, S Spain) is fed mainly by the infiltration of rainwater through carbonate outcrops and by a sinking stream entering the system through Villaluenga del Rosario shaft. Drainage occurs towards its SW border, through the springs of Cornicabra, Algarrobal (both permanent) and Garciago (overflow).
Systematic surface water and groundwater sampling was realized under different hydrodynamic conditions. Water samples were acidified with analytical grade HNO3- free of trace metals and soil samples were taken from the (clayey -terra rossa-) epikarst in Sierra de Ubrique and the (clayey – flysch) Villaluenga shaft catchment. The digestion of soil samples was realized following the U.S.EPA (1996) standards, and then, groundwater and digested soil samples were analyzed in the laboratory through the use of ICP-MS (Thermo Scientific ICAP- RQ). Turbidity and complementary parameters such as electrical conductivity, temperature as well as spring discharge have also been measured using a portable multiparameter and in-situ water height recording probes.
Among the analyzed elements, a reduced group of them (Li, B, Al, Mn, Ni, Ga, Sr and Ba) were reliably detected in groundwater samples. The soil content in these elements at the two soil sampling areas presents the same order of magnitude, except for Sr and Ba.
Maximum turbidity measurements in Algarrobal spring (Qmean=129 l/s, TurbMean=56 NTU) at each flooding event were ≈5 times higher than those obtained in Cornicabra spring (318 l/s, 9 NTU). Moreover, Algarrobal spring water shows a higher concentration and temporal variability of the studied elements than in Cornicabra. Algarrobal spring shows a significant statistical
correlation between Al and Mn with turbidity (pearson>0.91 and p>0.97 respectively) as well as happens at Cornicabra spring (p>0.87 and p>0.93). However, other elements (Li, Ga, Sr, Ba) show a linear correlation with turbidity from the threshold value of 20 NTU, so that this is more noteworthy at Algarrobal spring (0.83>p>0.71). This relationship is not that evident at Cornicabra spring (0.50>p>0) due to lower turbidity values. Finally, there are two elements (B, Ni) that show a weak correlation with turbidity (p<0), so that their transport in groundwater must be dominated by different mechanisms or a sum of them.
Thus, the influence of allogenic recharge is noticeable at both springs, as Al and Mn proceed from clay minerals hydrolysis coming from the shaft catchment. Nevertheless, it is more evident in Algarrobal spring, since it shows higher turbidity and a greater concentration of elements that enhance its mobility with the sediment load increase.
Whereas karst aquifers have peculiarities, they are also the source of a great deal of the world’s water supply and as such regular monitoring is essential for qualitative and quantitative analyses. Because karst is characterized by rapid transport of recharge through voids and conduits, rapid changes in water quality need to be well documented. New technologies allow for onsite measurements, but the installation of these sensors can provide substantial challenges when dealing with artesian situations.
In this presentation we will review the differences between artesian and non-artesian wells. We will explore the challenges that these wells present for the installation of sensors, and we will aim to offer some suggestions for possible solutions taking account of environmental and technical aspects.
We will explore the reason for wanting to monitor confined aquifers, the suitability of sensors with an accent on new technologies such as Nitrates, TSS and Turbidity and will discuss the requirement for long term stability of optical sensors.
New technology sensors which can provide vital water quality data quickly and reliably and whereas the installation of these sensors can be challenging, with good planning deployment is possible in artesian wells.
The Jura Mountains, known for their large karstic groundwater resources, are at higher altitudes locally covered by impermeable moraine deposits, favoring the establishment of superficial hydrosystems such as lakes and peatlands. The hydrological functioning of these peatlands remains poorly understood, while their role in sustaining biodiversity, storage of carbon as organic matter, and local water resources is of increasing interest, especially since the last 5 years during which the region suffered historical droughts.
To clarify the mechanisms constraining the transit and discharge of water in peatlands of the Jura Mountains and the possible interactions with the underlying karst aquifer, this study addresses the water and solute patterns in the Frasne-Forbonnet peatland (46.826 N, 6.1754 E; 850 m a.s.l). The site is a long-term observatory of the French Critical Zone Research Infrastructure (OZCAR). For this purpose, we combine hydro-meteorological monitoring (P, T, ET, Q from 2014 to 2019) with water geochemistry (major elements, pH, electrical conductivity, T, $δ^{18}OH_{2}O$ and $δ^{2}HH_{2}O$) at a monthly time scale during one hydrological year (Feb 2020-Feb 2021) and strontium isotope ratios ($^{87}Sr/^{86}Sr$) during low- (July. 2020) and high-flow (Feb. 2021) periods from a network of 20 piezometers along upstream-downstream and vertical gradients.
The relationships between seasonal to daily P inputs with Q and corresponding electrical conductivity suggest that the peatland is fed by a nested water system with 3 different water origins: (1) karst groundwater inputs across localized pockmarks at the bottom of the peatland, (2) seepage from surrounding wooded peatlands and (3) meteoric inputs. Karstic groundwater inputs seem to contribute to almost 50 % of the peatland outlet discharge at low flow.
This preliminary model is strengthened by the systematically depleted $δ^{18}O/δ^{2}H$ composition of the most mineralized bottom water layer, suggesting an origin from precipitations at higher altitudes and a subsequent transfer to the peatland via the regional karst aquifer. Consistently, $^{87}Sr/^{86}Sr$ and Sr concentrations are spatially variable and highlight that mineralization is mainly derived from carbonate dissolution and atmospheric inputs. Deep waters are characterized by a carbonate contribution to solutes of almost 100%, while the superficial waters are mainly fed by atmospheric wet deposition. This heterogeneous distribution of solute sources highlights the complexity of karst-peatland interactions.
Nevertheless, the interactions between the peatland and the karst aquifer are not solely characterized by karst inputs because the pluri-annual hydrological balance of the peatland is strongly positive. It suggests that the outputs by evapotranspiration and superficial discharge are complemented by outflows to the karst aquifer. This finding is consistent with the presence of a sinkhole next to the peatland and by similar features visible in ground penetrating radar images from previous studies.
These complex hydrogeological patterns revealed in the Frasne peatland suggest that it acts as an intermediate groundwater discharge-recharge site between the upper Jura Mountains and the downstream springs supplying large regional rivers (Ain, Loue). This function may be of importance regarding the impact of climate change over the regional resources, as the site could buffer extreme meteorological events and the resulting increase in discharge variability.
The microbial signature is a fundamental characteristic of karst groundwater resources, both with respect to the natural state and potential contamination. Karst springs are regularly affected by water quality derogation, particularly in the course of quick response to recharge events. Deeper resources with longer residence are often free of fecal contamination and turbidity problems, what puts them into the focus for future exploration strategies. Both aquifer types are however still few understood regarding their microbiological constitution. The present study assesses the microbial fingerprint of karst groundwater from shallow and deep aquifers, and discusses how this signature evolves from recharge to discharge at differing residence times.
Earlier work could show that the Total Cell Count (TCC) is an appropriate parameter for investigating and characterizing karst aquifers with respect to their microbiological state and vulnerability. Additional flow-cytometry parameters have the potential for marking further progress in this direction by specifying the microbiological status in terms of a cell fingerprint. The present study compares spring and well data in the same hydrogeological setting over a one-year period. Monthly sampling provided data for fecal indicators as well as flow-cytometry data (TCC, live/dead cells and LNA/HNA cells ratios) for differing hydrological conditions. The karst system is located in the Swiss Jura Mountains, and data were set in context with continuous discharge and physico-chemical measurements, which is the primary information to be checked with microbiological parameters occurrence and abundance.
While the spring results follow the hydrologic variations, the well data evidence microbial stability without correlation to periods of recharge. The high stability for deep groundwater could be linked to the increased residence time of deeper groundwater flow, approaching equilibrium with the natural biocenosis of the groundwater ecosystem. While spring TCC remains at least one order of magnitude above those of the well, other microbial parameters approach to each other for baseflow conditions.
The consideration of both spring and well monitoring assesses the end-members of karst groundwater dynamics and vulnerability. This (i) allows deepened insight into microbial fate and transport processes, (ii) explains the hydrogeological meaning of the new flow-cytometry parameters, and (iii) makes it possible to deduce general and specific conclusions from the microbiological signature according to the analytical data. The presented findings provide a concept for the evolution of diverse microbiological parameters with (residence) time, building the basis for the protection of potential future resources.
Contrary to most caves, in which temperature gradients between the surface and the cave drive convective compensating air currents, airflow inside barometric caves is induced by air pressure gradients between the surface and the cave, resulting from atmospheric pressure variations. Recent research in Wind Cave and Jewel Cave – two major barometric cave systems in the Black Hills of South Dakota (USA) – has shed new light on the dynamics of barometric cave airflow. Based on high-resolution long-term air pressure measurements from the surface and a variety of locations inside both caves, four differences between surface and cave pressure signals were identified: Compared to the free atmosphere, the pressure signals within Wind Cave and Jewel Cave showed (1) an absolute displacement due to different altitudes of the measuring sites, (2) a delay related to the travel time of the pressure wave to the measuring sites, (3) a smoothing effect, and (4) a dampening effect due to long response times of the caves to external pressure changes.
Based on these new findings, an iterative quantitative model was developed to predict air pressure inside barometric caves from external measurements. Therefore, each identified speleoclimatological process was translated into a mathematical operation: First, the absolute deviation caused by altitude differences between the cave locations and the surface is calculated and then added to the measured surface pressure. The delay of the pressure signal inside the cave is determined by cross-correlation analysis. Consequently, each value of the cave pressure signal can be assigned to a past value of the surface pressure signal. In order to model the smoothing effect on cave air pressure, the impact of the caves is mathematically described as a low pass filter, thus removing the high-frequency components of the surface pressure signal. Finally, the dampening effect is included in the model. By subtracting the mean of the measurement series from each pressure value, the resulting relative pressure fluctuates around zero and can therefore be "damped" by multiplication with a damping factor D determined by linear regression analysis. For evaluation, the pressure values calculated by the model are compared to the actual measurements of air pressure inside the cave. Correlation analysis yields an R² value of 0.996. Thus, the proposed model can predict 99.6 % of the measured air pressure inside the cave, suggesting that all significant processes modifying pressure waves inside Wind Cave and Jewel Cave have been recognized and adequately modeled. The remaining discrepancies can be attributed to measurement noise.
The proposed model also allows determining the direction and strength of the pressure gradient between the inside of barometric caves and the outside atmosphere, which then induces the characteristic compensating air currents. Considering the fundamental importance of underground airflow for almost all elements of speleoclimatology, including air temperature, humidity, and CO2 dynamics, as well as for management and conservation purposes, this model lays the groundwork for an improved understanding of numerous aspects of climate systems inside barometric caves.
In heterogeneous fissured-karst aquifers, much attention is often paid to the assessment of the vulnerability of drinking water springs based on delineating large scale catchments, or taking the protective effects of cover and epikarst in vulnerability mapping. Tracer tests have already become a standard tool in such approaches, helping in the definition of safeguard zones, even if tracer velocities must be carefully criticized before any extrapolation. Especially when a connection with the spring is already proved, including nearby karst features in inner protection zones is quite common. In fissured media, the anisotropy of fissures can also be taken into account. However, swallow holes with a complete absorption of the streams are not necessarily the rule in all karst areas and a fortiori in other fissured media. So, surface streams, some of which are not permanent, flowing very close to the springs, may have an insidious impact on the quality of drinking water, especially regarding point-source contaminations and spills. They may also produce chronical or temporary microbiological concerns, whose relationship with a turbidity indicator, or temperature, is not always clearly established. Such quality impacts are usually due to discrete, and often hidden, infiltrations of the streams through more or less enlarged fissures, which, despite some fillings, may allow high flow velocities on short distances towards the springs. Detecting tracers from such streams may become a challenge since dilution may be very high, breakthroughs are not necessarily so well structured in time as for classical tracer tests, and, especially, the flows are largely depending on the instantaneous stage of the river. Some hints are here presented, which focuses on the reliability of the detection of fluorescent tracers, their specificities of implementation in such streams, as well as the methodological aspects to locate or/and quantify stream leakages towards springs. Various cases are used to demonstrate the feasibility of such tests, as well as their limitations.
The development of karstic voids within a carbonate reservoir is governed by many controlling parameters that make the geometry of the networks difficult to predict. Karst specialists generally have to deal with this difficulty and evaluate the flow network through homogenized average parameters. However, the control and management of flows in karst hydrodynamic systems require more than a global approach, and the 3D modelling of cavities appears to be an increasingly essential step, prior to flow simulations for predictive purposes.
Among the parameters that control the spatial organization of vugs, conduits and caves, some belong to the geology-related reservoir heterogeneity, and constitute a inheritance for the karstogenesis process. The nature of the rock, its mineralogical, chemical and petrophysical properties, and the set of discontinuities that run through it, summarize the geological history of the reservoir. Other parameters are associated with the speleogenesis process, either through a control external to the system provided by the climate, or through parameters specific to the system such as the infiltration or the groundwater zonation. The complexity is increased when dissolution processes are superimposed, from very early to late contemporary phases.
Building models capable of delivering a realistic architecture of conduit networks consistently with all knowledge seems a priori a hopeless challenge. Modelling methods exist, but they are generally based either on geostatistical reproduction of observed structures without testing the chain of causality, or on techniques that give prominence to the chemistry and kinetics of the CaCO3+H2O+CO2 system and that partially ignore the complexity of the medium in which the reaction takes place.
The approach presented in this paper links the various geological and hydrogeological parameters at the origin of the current complexity of karst reservoirs. The combination of several simulation tools allows the delivery of models that ensure spatial and temporal consistency between processes.
The forward stratigraphic simulation of the depositional process first generates the sedimentary facies, with the mechanical and petrophysical properties of the medium that partly controls the distribution of fractures in the reservoir. In the case of early dissolution affecting the medium, the potentially affected zones are identified. The stratigraphic architecture resulting from this simulation generates remarkable stratigraphic surfaces likely to play an important role during karstification. Then, fractures are modelled by the generation of Boolean objects according to a process that respects the mechanical properties of the medium and the tectonic deformations. Finally, the speleogenesis is simulated by means of a cellular automaton that uses the properties derived from the previous simulations as a foundation on which the dissolution of the medium occurs. This process can be repeated with different hydrogeological parameters when successive phases of karstification have affected the reservoir.
Applied to the Urgonian formation, a Barremo-Aptian reservoir in south-eastern France, this disruptive workflow has allowed the generation of a karst architecture consistent with matrix properties, fracture network, and all available knowledge. It allows a certain confidence in the prediction of the unexplored zones of the Fontaine de Vaucluse groundwater, which will have to be validated by hydraulic simulations.
Several global groundwater models have been developed to estimate water availability or water stress under human impacts and climate change. However, a differentiation of karst aquifers in these models is absent. Due to the special features of karst, such as conduit networks, the laminar flow approach used in global models cannot well represent the fast flow processes in karst regions. At large scale, it is difficult to obtain karst conduit distributions and their properties. Therefore, how to conceptualize the flow in conduits becomes crucial for regional to global scale modeling. In this study, we present a grid-based regional karst groundwater model. We model matrix flow with a laminar flow approach and model conduit flow using the concept of discharge recession and lumped process representation. The groundwater flow direction in the karst aquifer is computed from matrix groundwater head simulations. The fast flow in conduits follows this groundwater flow direction and discharges to springs and streams in the karst regions. Non-karstic areas in our simulation domain are modeled using a lumped catchment model. We test our approach in the upper Danube and upper Neckar catchments. Our results show that the hydrodynamics of the daily streamflow can be simulated acceptably with our uncalibrated model. The correlation coefficient between the simulated specific discharge and the observed spring discharge can be larger than 0.65, demonstrating a good performance on simulating karst processes. It also indicates that our approach generating groundwater discharge points taking account of river network densities to represent karst springs is plausible given no available information about karst spring locations and properties. The groundwater head change in the karst matrix is sensitive to the diffusive recharge. Our regional simulation underestimates few large events, suggesting a surface runoff routine should be improved to capture the very fast flow component. Future work using trace experiments can be helpful for evaluating the groundwater flow direction and for conceptualizing the fast flow paths. Our regional karst groundwater model provides a feasible way to include karst processes in the large scale groundwater modeling.
The STREAM research project (2017-2021) aims at leveraging expertise in karst dye tracing, hydrogeology, hardware/software engineering and instrumentation to develop a compact and easy-to-use field fluorometer for hydrogeology with focus to karst and complex environments.
First prototype of this tool was presented at the 2018 Eurokarst conference in Neuchatel. Since those developments, a multidisciplinary program gathering hydrogeologists, cavers and engineers successfully develops, tests and validates a professional tool. The STREAM field fluorometer is a compact (1,25 kg, 23x6cm) standalone probe, including optical probe, internal battery and data logger in the same casing. The tool is 100% submersible since it was designed to be easily installed in challenging environments, possibly flooded or under pressure (pipes networks…), field test validates a -800m depth rating. The fluorometer was successfully tested in field conditions with 3 fluorescent dyes (sodium fluorescein, sulforhodamin B, amino-g acid), in addition to turbidity and temperature (all parameters can be measured at each timestep). A user-interface is embedded so no additional software is required for configuration, data visualisation/download or calibration.
Data from the three last years of the STREAM project are presented, mainly in Belgium karst systems. The focus will be on the development choices, field performances of the instrument, its advantages and limitations for various use-cases. Future challenges and perspectives for research projects based on the STREAM fluorometer will be addressed.
Karst catchments are highly sensitive to climate change and human disturbances. In recent decades, karst mountain areas exploited for livestock and agriculture have been gradually abandoned. These abandoned lands are progressively recolonised by forest (Houet et al., 2012). These changes in land use result in environmental impacts poorly investigated due to the scarcity of long-term historical data and the lack of suitable methods. In this context, the wooded Baget catchment in the Pyrenees Mountains was selected to develop a method for rebuilding the evolution of the landscape in connection with the long-term hydrological and hydrochemical survey since 1979 (Binet et al., 2020, Ulloa-Cedamanos et al., 2020). The chosen method is based on two multi-temporal image sources (aerial photographs from 1942 to 1989 and satellite images from 2001 to 2019) and heterogeneous data (DEM, forest inventory and Google Earth). The landscape classification included three main groups: closed (forested, with broadleaf or coniferous trees), semi-open (loose broadleaf coppice) and open environment (grasslands and fern heaths). The landcover evolution showed a regular increase of the closed environment by 31% over the whole period (average increase of +0.4% per year). The impact of the landscape change was discussed together with the observed increase in air temperature (0.03°C per year, p<0.001) in this remote catchment over the last 40 years (Ulloa-Cedamanos et al., 2020). These changes were compared to the trends of streamwater discharge and water quality, suggesting a set of key processes influencing these trends, particularly for dissolved elements derived from rock weathering (calcium, magnesium and bicarbonate). The afforestation (closure of the landscape) promoted a higher production of soil organic matter, which increased pCO2 in soils and promoted carbonate dissolution. In addition, the increase in air temperature enhanced tree growth, tree water demand and litter decomposition. This work revealed the potential of reconstructing landscape evolution to better understand the long-term hydrochemical trends and to highlight the major processes in the critical zone influencing these changes.
References :
Binet S., Probst J. L., Batiot C., Seidel J. L., Emblanch C., Peyraube N., Charlier J.-B., Bakalowicz M. and Probst A. (2020) Global warming and acid atmospheric deposition impacts on carbonate dissolution and CO2 fluxes in French karst hydrosystems: Evidence from hydrochemical monitoring in recent decades. Geochim Cosmochim Acta 270, 184–200.
Houet T., Vacquié L., Vidal F. and Galop D. (2012) Caractérisation de la fermeture des paysages dans les Pyrénées depuis les années 1940. Application sur le Haut-Vicdessos. Sud-Ouest Eur Press Univ du Mirail, 41–56.
Ulloa-Cedamanos F., Probst J. L., Binet S., Camboulive T., Payre-Suc V., Pautot C., Bakalowicz M., Beranger S. and Probst A. (2020) A forty-year karstic critical zone survey (Baget catchment, Pyrenees-France): Lithologic and hydroclimatic controls on seasonal and inter- annual variations of stream water chemical composition, pCO2, and carbonate equilibrium. Water 12, 1227.
Electrical Resistivity Tomography (ERT) experiments are widely used for characterizing the natural environment because they easily provide a large number of data on a large extent and at various depths. These experiments are usually interpreted with forward models and inversion strategies that are well suited for equivalent porous medium representations. Dealing with fractured rocks requires using specific forward models, which are able to simulate the propagation of electric current flow in fracture networks that are embedded in conductive matrix-rock. Although such models have been recently developed, their integration into inversion strategies, and the definition of the inversion framework are still challenging. Here, we will start by showing how simple cases such as a single horizontal and vertical fracture can be handled by assuming the fracture position as known and inverting the fracture connectivity. We will then extend this to a regular network of fracture segments and will analyze various metrics of the inversion procedure, such as the convergence rate, resolution and depth of investigation. We will also discuss how defining prior information to improve the stochastic inversion strategy and future coupling with other characterization methods.
Despite recent advances in modelling the evolution of karst systems, "old" 2D models of fracture network evolution can still provide deep, intuitive insights into the mechanisms of early speleogenesis in different environments. We demonstrate the use of these models to investigate two distinct mechanisms: (1) breakdown of dissolution front in fracture networks and (2) mixing zone instability in an early hypogene setting. (1) When a set of identical, parallel fractures develop under constant head conditions, all fractures grow at the same rate and break through simultaneously. However, if the fractures are interconnected into a perfect rectangular grid, the initial uniform dissolution front collapses and only a few fractures continue to evolve until their breakthrough. We provide a model-based intuitive explanation of how this occurs and discuss the importance of this mechanism for early karstification and the use of different modelling approaches. (2) In a second example, we present the evolution of a fracture network driven by mixing corrosion in an unconfined aquifer caused by an inflow of deep-seated water with high pCO2. In a network with statistically distributed initial apertures, slight differences in the mixing ratio cause differences in dissolution rate within the mixing zone. This makes the mixing fringe unstable, where flow, mixing and dissolution is focused to a narrow region from which the aggressive solution is re-injected into the surrounding network. This leads to a specific porosity pattern and geometry of the resulting conduit.
The Conduit Flow Process (CFP) extends the numerical groundwater flow code MODFLOW-2005, accounting for flow in discrete conduits of karst aquifers. It allows for the simulation of laminar or turbulent flow in one-dimensional discrete pipe networks coupled to a matrix continuum. CFP relies on specific input data to define the discrete conduit network. We developed Python-based pre- and post-processing routines to generate model input files and to obtain and process results such as state variables of nodes, conduits and surrounding matrix (e.g., hydraulic heads and conduit flow rate). Additionally, a visualization utility was developed, which enables a 3D dynamic representation of the nodes, conduits and layer elevations as well as simulation results. The developed method enables an effective interconnection between (stochastic) karst network simulators – or available data on conduit networks – and MODFLOW-CFP. The conduit network structure is a major factor controlling groundwater flow in karst aquifers. Thus, the influence of different structures needs to be explored in the model calibration to identify an adequate system representation. The package offers the possibility of automatically altering the network structure and parametrization to get effective system understanding and to lower the structural uncertainty. Furthermore, the package now allows for an effective creation of MODFLOW-CFP models in the Python framework of FloPy, ensuring a reproducible and script-based model development.
The package utilizes MODFLOW layer elevation information on the one hand and conduit network node elevation on the other hand to generate pipe connections. It is possible to create networks having multiple vertically connected node planes independently of MODFLOW layers. The input data format to the package is an ASCII file containing (1) general model discretization, (2) uniform or spatially distributed layer elevation and (3) spatially distributed node elevation information for each node plane. Spatially distributed information is given in matrix or two-dimensional array format.
We demonstrate the functionalities of the package using several synthetic examples. It is shown how conduit network structures can be created and subsequently used as an input to the Python package. It is also shown how the package can be used within the FloPy framework and how script-based karst hydrological modelling, including pre- and post-processing, can be conducted. The different visualization capabilities for pre- and post-processing are demonstrated as well.
Karst aquifers are characterized by the presence of dissolution features in bedrock that lead to high connectivity between the surface and subsurface, facilitating the introduction of surface contaminants into these systems. Microplastics (plastics < 5 mm) are emerging contaminants that are ubiquitous, having been detected in virtually all aquatic environments, including surface waters, the deep ocean, and arctic sea ice. Microplastics often degrade slowly and are highly mobile, allowing them to travel long distances and be easily ingested by wildlife. Microplastic research has been historically focused on marine and surface freshwater environments, with groundwater systems remaining understudied. Thus, our study aims to identify microplastic sources and transport mechanisms through a cave stream system. To do so, we monitored in situ water quality (e.g., temperature, specific conductivity, pH) and discharge from February 2020 to February 2021 for a stream issuing from a cave hosted in Mississippian St. Louis Limestone (Cliff Cave; St. Louis, Missouri, USA). We also collected water samples under a range of flow conditions, both weekly during the entire study period and at high frequency during floods (with four sampled flood events in total). Samples were analyzed for microplastic content and characteristics as well as geochemical parameters to help identify microplastic sourcing (e.g., suspended sediment concentration (SSC), ion chemistry, and δ18O and δ2H for hydrograph separations). Microplastics were found in almost all samples (91.5 %), with fiber as the dominant morphology (85.7 %) and clear as the most common color (54.6 %). Median microplastic concentrations were 3.1 microplastics/L in our weekly samples, but reached up to 81.3 microplastics/L during flooding. Total microplastic concentrations had significant, positive correlations with discharge and SSC (R2 > 0.14; p < 0.05) and significant, negative correlations with specific conductivity and pH (R2 > 0.17; p < 0.05). These trends indicate higher microplastic transport in the cave stream when dilute, sediment-rich event water enters the karst, compared to minimal microplastic conveyance during low flow. Antecedent moisture conditions may also play a role in microplastic transport through karst, because floods that occurred after dry periods transported higher microplastic loads compared to floods with similar peak discharge values that followed wetter conditions. Our study gives new insight into how microplastic contamination is transported to and through karst systems, which will help inform ecosystem and water resource protection efforts, as well as debris mitigation strategies.
The degree of belief in a hydrological model is mainly dependent on how well the simulations are realistically predicted, which is mostly achieved by ‘successful’ model calibration. Since the performance metric – a statistical or a mathematical measure reflecting the degree of agreement between the predictions and observed values – renders the indication of how well and to what degree the system of interest is represented, the model calibration requires the selection of an appropriate performance metric to verify the prediction accuracy. However, as there is not a general procedure or a rule for the selection of performance metrics, it is ultimately the modeller’s subjective decision. This study explores the contribution of the discharge young water fraction, which is derived from stable water isotopes, to select a hydrologically appropriate metric for the model calibration scheme. For our analysis, we examine metric combinations over the multi-criteria calibration in which model performance on discharge and model performance on discharge young water age make the best couple such that the optimized discharge criterion cannot improve without discharge-age metric worsening. For that purpose, we served 16 performance metrics for the assessment of model performance on discharge - including deterministic metrics and hydrological signatures – while the model is simultaneously constrained by the information of discharge young water age. Therefore, the trade-off between both metrics does not only cover the process-based discharge age information, yet also satisfies the minimization of model bias. Our finding indicates that discharge young water fraction allows selecting a hydrologically -more- relevant performance metric, thereby ensuring the physically more realistic parameter sets. The results entail that discharge-age information not only reduces the modeller’s subjectivity on the metric selection, which in turn leads to an improvement in the model parameterization, but also extracts hydrologically informative model predictions.
Understanding recharge processes is essential to improve the management of water resources. In the context of karst systems, soil and epikarst have a key influence on these processes. But there are only a few studies that directly investigate the shallow zone of the karst. In this study, a dataset including soil moisture, precipitation, and discharge measurements collected at a karst area in Southwest Germany is used to investigate infiltration through the epikarst zone. An event-based method was developed to study the link between soil moisture and recharge. The measurements were conducted hourly for 9 years at 15 soil profiles across a recharge area distributed including both woodland and grassland areas. Rainfall, soil moisture, and recharge events were extracted automatically and then linked to each other. We found a high spearman’s rank correlation between the mean soil moisture during events and the induced recharge. Then, a unit-gradient soil model to predict recharge from soil moisture measurements was performed showing acceptable results at the scale of the system and the temporality of the event. To show its applicability across various environmental settings, we currently apply our method at five other karst sites worldwide spread over five distinct climates and varying land covers. If the test is successful, our approach will allow the direct investigation of surface and subsurface hydrodynamic processes to a wide range of karst systems. As soil moisture measurements on the surface are relatively easy to perform, the new approach is potentially very useful in karst regions that currently do not have any information about groundwater recharge. Obtaining quantitative estimation of recharge using soil moisture will also help to better characterize recharge processes through the soil and therefore improve karst hydrological modeling and the management of water resources ultimately.
Fluorescence spectroscopy approaches as Tryptophan-Like-Fluorescence (TLF) have been presented as a powerful tool to easily detect bacteriological contamination in groundwater used for drinking water supply in rural areas, since bacteria are able to synthetize L-Tryptophan, which is an essential amino acid for biological processes. Given that the accuracy and reliability of this method has been little tested in karst aquifers, this work aims to show the preliminary results after 10 months of continuous record of TLF as a potential early warning parameter at Sierra de Ubrique Jurassic karst aquifer (S Spain).
Allogenic recharge in this binary karst system supposes a major threat to the aquifer water quality since a well-developed karst network exists and human activities (cattle, cheese factories and waste water of a little village of 500 inhabitant) inventoried in the feeding area, certainly increase groundwater vulnerability. Aquifer discharge is mainly produced through three karst springs located 6 km downstream of the sinking point.
During flooding conditions, episodic high turbidity levels (and bacteria detection) hinder the total exploitation of groundwater. TLF has been implemented as a potential early warning parameter to prevent polluted groundwater to reach drinking water capture points. Continuous record of TLF was realized with two GGUN-FL30 field fluorometer (Albilia Co) (holding specific channels for excitation/emission around 280 nm/350 nm) installed at the permanent outlets (Cornicabra and Algarrobal springs) with a sampling interval of 15 minutes. Complementary parameters such as electrical conductivity, temperature and turbidity as well as spring discharge have been measured and Total Coliforms and E. Coli have been analyzed by hand sampling.
The different TLF maximum intensity: ≈120 eq. ppb at Algarrobal spring and ≈40 at Cornicabra evidences a higher contaminant load at the first spring, which also makes sense with maximum turbidity records (≈350 NTU and 60 NTU respectively). An apparent correlation has been found between biological activity indicators (Total coliforms, E. Coli) and the equivalent concentration of TLF at Algarrobal spring (p=0.75), however, Cornicabra spring does not show a good correlation between them (p=0.26).
Some of the tryptophan peaks with significant concentration do not coincide in time with rain episodes and they are not correlated either with turbidity or the presence of bacteria, which increases the uncertainty of the measurement as early warning parameter. The obtained chemographs in each spring suggests the complexity in obtaining reliable continuous measurements as this technique shows multiple limitations (such as the decay of light intensity or interference with turbidity, iron precipitates and other compounds that disturb the signal intensity) that arose during the study period. Thus, the multiple potential origins of tryptophan (cheese whey, bacterial activity in soil or karst conduits…) together with possible biochemical reactions in the system of karst conduits relegate TLF to a function as a biological risk indicator rather than an early warning parameter.
Because karst aquifers provide about a quarter of the world's drinking water and are special habitats, the quality of karst waters is of paramount importance. While it is known that water quality generally deteriorates during recharge events, legal guidelines lack a concrete framework for adequate monitoring. Therefore, this study fills the identified gap by proposing more specific recommendations for water quality monitoring. These are based on detailed hydrochemical and microbiological sampling of natural tracers during two flood pulses. The study was conducted in a hydrodynamically complex karst environment in the Unica springs catchment in southwestern Slovenia. It was found that the responses of the studied springs are characterised by the different recharge dynamics of the hydrogeological sub-catchments that determine their behaviour. The results show that water quality fluctuates the most at the beginning of the flood pulse. This is mainly due to the effects of sinking rivers' quality, which affect the springs with a delay of about one day. Therefore, the appropriate locations, frequency, and duration of monitoring were determined. Monitoring the quality of sinking rivers should serve as a preventive measure. Relevant monitoring parameters such as bacteria, turbidity, Cl, EC and Ca/Mg ratio were identified as indicators of water quality deterioration. Since turbidity and EC are relatively simple and easy to measure in situ, they allow a quick and inexpensive assessment of the situation. The results obtained could help to detect contamination in time. Potential drainage areas as sources of poor quality water could be identified through contamination risk assessment. Although the proposed guidelines are site-specific, they could be applied to other comparable karst areas. They could be used to develop recommendations for early warning mechanisms for contamination and protection of karst water resources.
Acknowledgement: This study was conducted within the context of the projects No. L7-2630, J2-1743, No. NK-0002, and the programme Karst Research, No. P6-0119, all financially supported by the Slovenian Research Agency.
Karst aquifers are complex systems with high a heterogeneity and anisotropy. The duality of karst is highlighted by several processes such as recharge, in-filtration, storage, and flow. Given that numerical models based on Darcy’s law are limited to simulate turbulent conditions in complex conduits, an approach to simulate the effect of conduits, or preferential flow layers, is necessary for karst. The Conduit Flow Process (CFP), developed by the U.S. Geological Survey (USGS), helps to simulate this duality in flow commonly found in karst aquifers. In this work, the CFP was applied in the Yucatan karst, categorized as a well-developed karst with conduits of considerable diameter in the subsurface. Unfortunately, there is a lack of data regarding the three-dimensional location and connectivity of conduits and underwater springs. For this reason, the CFP in mode 1 was applied evaluating different conduits arrangements based on indirect data from gravimetry and geophysics. The Merida Metropolitan Area (MMA), a densely populated region in the Mexican state of Yucatan, was selected as the area of interest for this study. Piezometric data from 48 monitoring wells, for the period 1996-2004, were utilized as the basis for model inversion; two measurements per year (representing dry and wet seasons) were utilized for the temporal discretization of the model. Similarly, precipitation and evaporation data from 14 climatic stations were included as stressors of the system for the same period. Results were compared with those from previous groundwater flow models that have been applied in the same area utilizing an equivalent porous media (EPM) approach and the CFP in Mode 2 (interbedded preferential flow layer). Results from the application of the CFP are expected to provide important insights regarding groundwater flow in this karst region. From this work, important considerations for further studies regarding recharge-discharge processes in Yucatan are highlighted.
Karst aquifers are known to be highly heterogeneous with a hierarchically organized network of conduit and matrix drainages – each exhibiting different porosity and complex hydraulic interplay. Understanding the complex interactions of the different drainage systems is very vital for the management of karst aquifers. The effectiveness of the hydraulic connectivity between the conduit and matrix controls the rate of groundwater movement into and out of the aquifer after storms events and during base-flow conditions. Recession analysis of karst spring hydrographs is therefore a vital technique for investigating the hydraulic functioning of the different drainage domains and estimating groundwater flow parameters that control the storage-outflow processes. Recession analysis of karst spring hydrographs involves decomposing recession curve into multiple segments. Often, spring hydrograph separation is done through a subjective graphical approach and recession parameters are estimated either by analysing (1) individual recession curve or (2) master recession curve constructed from all recession events. While the master recession curve approach is widely used in karst hydrology, parameters derived from this approach do not provide a better reflection of the variable hydrological response and heterogeneity of the karst system. The approach is considerably biased towards long recession events and does not capture the average behavior of the system. In our study, we developed a new automated approach for separating conduit and matrix recession components as well as a robust parameter estimation method. We introduced a recession event-dependent parameter called initial matrix fraction, f, accounting for varying matrix storage conditions at the beginning of recession. Afterwards, a two-linear (Maillet) reservoir model was used to simulate the storage-discharge behavior of the conduit and matrix reservoirs during recession. In addition to that, and for the first time in a karst recession study, we also introduced a Monte Carlo framework that allowed for a robust parameter estimation and quantification of associated uncertainties. The approach was tested on a sample of karst spring system in different climate zones. With KGE values > 0.7 produced by the best 20% of the Monte Carlo parameters sample, our results showed (1) that karst aquifer drainage can be adequately represented by a two-linear drainage model, (2) that the initial matrix fraction, f, is related to the aquifer’s old water fraction indicating matrix contribution to total spring flow and (3) distinct optimal values and narrower ranges of posterior distribution of the conduit parameter while the matrix parameter showed less sensitivity with broader ranges. With this new approach, we were able to produce an effective range of the conduit and matrix recession parameters, which reflect a better dynamic description of the aquifer variable responses and heterogeneity of karst system. Our new approach is completely automated thereby suitable for large sample of spring hydrographs analysis. It is also very flexible and can be applied using different types of storage-discharge models. We believe that the robustness of this approach with attention to uncertainties quantification provides useful information, which is greatly important for groundwater modelling and the general management of karst water resources.
Karst environments around the globe host precious freshwater resources for humans, as one quarter of the global population is completely or partially dependent for drinking water from karst aquifers (Hartmann et al., 2014; Chen et al., 2017a,b; Stevanovic, 2018; Goldscheider et al., 2020). Given their physical heterogeneity and anisotropy, they need specific adaptation in the investigating techniques, which need to be different from the classical hydrogeological methods applied for porous or fractured aquifers. Although it is certainly necessary to collect a variety of geological field data of the investigated areas at the surface, monitoring actions at high temporal resolutions (ideally, continuous monitoring) both at the surface and directly within the subsoil are required to properly characterize the dynamic behavior and variability of karst systems. In this sense, presence of caves, and the possibility to directly enter karst systems, offers an unvaluable opportunity to cavers and scientists. Understanding the way of functioning of karst aquifers is not a simple matter, also because their characterization through direct data collection in such an environment results often difficult. Anyway, the building of a karst conceptual model, using all available data, has to be considered the primary and necessary step to understand the hydrodynamics and hydrology of karst systems. This is the first, needed, step for the subsequent implementation of numerical models aimed at predicting the modalities of flow and transport of groundwater resources in karst.
Taking advantage from the possibility to explore and reach directly the water table in two cave systems of Apulia (Southern Italy), we present here methods to characterize the karst groundwater at these two study areas, coupling data derived from classical geological survey at the surface and direct measurements in cave.
In detail, the activities carried out at the surface are structural surveys, production of a karst geomorphological map, and collection and interpretation of available ancillary data; on the other hand, below the ground, the cave was equipped with probes to monitor groundwater parameters and cave micro-climate, and explorations of both dry and flooded channels was performed, together with collection of rock samples for thin section analysis. Water samples were taken periodically from wells, and directly from the unsaturated and saturated zones within the karst systems, in order to carry out chemico-physical, micro-biological, isotope and ecological analyses. The final aim of this research, still ongoing, is to implement numerical models where the combined use of surface and subsurface data guarantees to obtain more realistic model outputs, and to contribute to improving the knowledge about the hydrogeological dynamics of the Apulian karst.
REFERENCES
Chen Z., Auler A.S., Bakalowicz M., Drew D., Griger F., Hartmann J., Jiang G., Moosdorf N., Richts A., Stevanović Z., Veni G. and Goldscheider N. (2017a). The world karst aquifer mapping project: concept, mapping procedure and map of Europe. Hydrogeology Journal, 25 (3), 771–785.
Chen Z., Goldscheider N., Auler A.S., Bakalowicz M., Broda S., Drew D., Hartmann J., Jiang G., Moosdorf N., Richts A., Stevanović Z., Veni G., Dumont A., Aureli A., Clos P. and Krombholz M. (2017b). World Karst Aquifer Map (WHYMAP WOKAM). BGR, IAH, KIT, UNESCO.
Goldscheider N., Chen Z., Auler A.S., Bakalowicz M., Broda S., Drew D., Hartmann J., Jiang G., Moosdorf N., Stevanović Z. and Veni G. (2020). Global distribution of carbonate rocks and karst water resources. Hydrogeology Journal 28, 1661–1677.
Hartmann A., Goldscheider N., Wagener T., Lange J. and Weiler M. (2014). Karst water resources in a changing world: Review of hydrological modeling approaches. Rev. Geophys., 52, 218–242, doi:10.1002/2013RG000443.
Stevanović Z. (2018). Global distribution and use of water from karst aquifers. In: Parise M., Gabrovsek F., Kaufmann G. and Ravbar N. (Eds.), Advances in Karst Research: Theory, Fieldwork and Applications. Geological Society of London, sp. publ. 466, 217–236. https://doi.org/10.1144/SP466.17.
Karst groundwaters in agricultural areas are highly vulnerable to nitrate contamination due to their thin and mostly freely drained soil cover. Nitrates are highly water-soluble and therefore, in particular during rainfall events, readily transferred from the soil to the aquifer. Within the aquifer they reach the system outlet within hours to days without purification. Nitrate concentrations in karst springs are therefore known to vary rapidly with discharge, which makes high-frequency monitoring indispensable for an adequate characterization of the system.
The karst rivers of the Jura Mountains are renowned for their excellent water quality and for fly fishing. This idyllic setting has been tarnished since 2009 by a series of massive fish diseases during summer low flow, demonstrating that water quality is actually worse than suggested. Since then various studies have shown that the diseases have multiple origins, but that one of the key factors is nitrate-driven proliferation of macroalgae and eutrophication. The main nitrate source is diary farming for cheese production in the infiltration zone of the karst aquifers feeding the rivers.
The nitrate dataset for rivers and karst springs available in 2010 was mainly based on daily sampling by a river water abstraction site for the drinking water supply of the city of Besançon and on irregular manual sampling. High frequency monitoring began in 2015 simultaneously for the Loue river (Quarstic project by BRGM, Charlier et al. 2018) and for selected karst springs of the Jurassic Karst observatory (french SNO Karst network). We will present here high-frequency data acquired since July 2015 by one of the monitoring stations of the Jurassic Karst observatory located in the village of Lods in the Loue valley.
The dataset was acquired with an automatic UV-vis probe (s::can spectro::lyser, 30 minute time-step), completed with samples recovered with an automatic sampler every 4 days. The data are characterized by annual cycles with relatively low concentrations of about 10 mg/L of NO3 in spring and summer and with peaks of up to 40 mg/L of NO3 during storm events in autumn and winter. The highest concentrations typically occur at the first storm event in autumn, followed by progressively decreasing peaks for the subsequent events. In our presentation we will compare this nitrate record with other parameters monitored concomitantly (water level, temperature, electrical conductivity, turbidity) in order to elucidate in detail the dynamics of nitrate in this karst aquifer.
Karst aquifers are fundamental in the water supply of European countries, where outcrops of carbonate rocks are very common, providing abundant groundwater resources, available for human use and environment preservation. The need to update the groundwater budget, at local up to at the Mediterranean scale, is due to the increase of human withdrawals but also to the effect of climate change, which can modify the recharge and consequently the spring discharge in amount and regimen.
In this framework, the KARMA project granted under the PRIMA call (Horizon 2020 programme) of the European Union, is dealing with the hydrogeological understanding and sustainable management of karst water resources to obtain for the entire Mediterranean area valuable information on recharge, groundwater vulnerability and groundwater-dependent ecosystems.
The Gran Sasso karst aquifer, selected as representative study area for Italy, is characterised both by high percentage of withdrawals for drinking purposes and significant interaction between groundwater and underground works. The peculiarity of the Gran Sasso aquifer is also due to the existence of a National Park, which testifies the double need to satisfy the human requirements without compromising the environment and the related ecosystem services.
For this reason, the water balance of the Gran Sasso aquifer has been evaluated to improve the knowledge about its recharge rate with respect to the previously collected data. The recharge evaluation has been carried out considering the 2001-2020 monitoring period and comparing three different methods: the Turc and Aplis methods, on annual scale, and the Thornthwaite method, on monthly scale, territorially distributed by 100x100m cells on GIS basis.
The total recharge considers not only rainfall but also the contribution of snow melting on infiltration, based on the real last ten years data and assuming a fixed 15% of the recharge for previous periods, with application to each of three different methods.
The results show similar mean recharge values in 2001-2020 for all methods, corresponding to 19.9, 18.5 and 19.4 m3/s respectively from Turc, Thornthwaite and Aplis methods. A significant contribution to recharge from snowmelt has been confirmed (3.2 m3/s included in the above-mentioned values). These values can be considered reliable with respect to real discharge of the regional aquifer.
To validate the recharge calculated values, the discharge data of the main springs have been also collected and analysed. The calculated recharge by different methods shows a general underestimation of the total spring discharge, highlighting that aquifer discharge is higher than annual stored recharge. This is probably due to a “memory – effect” which modulates the response of the aquifer to meteoric recharge.
The obtained results can be used to provide updated information to the drinking water companies, at the scale of single spring catchments, for a suitable management of the available resource, by regulating the flow rates provided for drinking supply but for other uses too in an optimal and sustainable way. The evaluations of the recharge-discharge contributions are useful also to estimate the potential impacts of climate change effect on groundwater resources, to be possibly extended at the Mediterranean scale.
Research conducted in the KARMA project (granted by European Commission under PRIMA program)
When a karst aquifer is located at the edge of a sedimentary basin, both autogenic and allogenic recharge can feed the system. when it occurs, outlet springs water origins can be difficult to assess which leads to issues in water resource management. To better understand these origins and the variations in the hydrodynamic and geochemistry of the outlets of a large and complex karst aquifer, this study presents a multi-proxy approach. It combines origins and groundwater flow path identification with geochemistry analysis, better hydrodynamic connections understanding with the use of crosscorrelation analysis, and assessment of respective contributions of the different origins (autogenic and allogenic) in outlet springs using a mixing model based on the geochemistry data. These methods are applied on the Ouysse karst system. It is a large watershed (650km²) located in Western France and is exploited for its drinking water resource. The upstream part of the karst aquifer is fed by three main water origins: (i) diffuse infiltration on the karstic limestone bedrock; (ii) sinking stream coming from igneous-metamorphic rocks runoff; (iii) sinking stream fed by highly mineralized evaporite spring. This study allowed to understand how these various origins feeds the different outlets in different hydrodynamics conditions. The mixing model at the main outlet shows about (i) 70-80% of its water coming from direct karst infiltration; (ii) 7-20% of water from igneous sinking streams; (iii) 1-20% of evaporite water. These results are currently used to build a rainfall - discharge analysis and modelling with karstmod in order to forecast variations for drinking water supply management.
Karst aquifers and springs are relied as the main drinking water supply in many countries. It is crucial to understand how the karst water resources may vary under the ongoing climate change. Many studies have attempted to quantify how the karst aquifers have and may respond to global warming. However, it is still not clearly understood how the karst aquifers have been influenced by the historic long-term climate variations. In this study we aim to contribute to this understanding by quantifying the response of the karst discharge to climate variations in the past 70 years at the Blautopf spring.
As the second largest and one of the most legendary karst springs in Germany, the Blautopf spring exhibits high economic and environmental values for protection. The long-term records show that the daily discharge rate at the Blautopf spring varies significantly between 0.3 m3/s to 32.5 m3/s, with a mean of 2.3 m3/s. Due to the groundwater memory, the aquifer responds to the precipitation events with time lags. It hence leads to the difficulty to predict the timing and intensity of the peak flow of the karst spring which is often desired for water management and supply. This task becomes even more challenging under climate change during which the precipitation pattern becomes more variable and the increasing temperature influences not only evapotranspiration but also the timing of snow melt.
Analysing individual karst discharge hydrographs potentially allows for a data-driven understanding of the karst spring behaviour under the historic climate variations. In this study we firstly adopted the statistical measures to quantify the karst aquifer memory to the precipitation. We then determined the trends of the mean monthly, seasonal, and annual discharge rate and quantified the relations between the discharge changes and meteorological shifts. In addition, a lumped karst discharge model was integrated with a snow melt module to analyse the response of the internal storages and fluxes of the karst system to the climate variations. Results show that the seasonal karst discharge exhibits a significant declining trend in spring but increased in winter as a response to global warming. The frequency, intensity, and time of emergence of the high discharge events have also changed due to climate variations at this site. By quantifying the impact of the long-term climate variability on karst discharge, this study contributes to an improved understanding of the karst aquifers response to climate change and future predictions.
Fracture networks control the permeability of many reservoirs. Since the fracture patterns in situ are therefore difficult to study in detail, field analogues are very important for understanding their fracture-related permeability and spatial organizations. Modelling of natural fracture in reservoirs requires, as input data, the results of a previous detailed and accurate spatial survey of the 3D fracture network. Here, we report the results of a systematic study of the fracture systems exposed in different outcrops of carbonate Formations in the Umbria-Marche Apennines, in Central Italy to defining a conceptual model of structural and hydraulic properties of the main basal aquifers. In the region, the main aquifers are located in the carbonate banks in the core of the anticline, and the groundwater supplies the springs located in the most important valleys or the main fault zones with an average base flow discharge of 22 l/sec/km2. The recharge occurs in the karst carbonate areas. Aquicludes are represented by marly layers that isolate local suspended aquifers. Groundwater flow in the transfer zone is controlled by karstic conducts and fissures. The hydrodynamics of larger springs is regulated mainly by the base flow, while in some cases quick flows mark the transfer zone. The geochemical characteristics of the groundwater are mainly carbonated with significant enrichment of sulfate in the base levels. The measured attitude and other structural parameters of more than 4000 fractures and bedding planes have been acquired in several key sites. Result two main fracture sets, one strike between SW-NE (dip-direction of N115) and other NNE-SSW (dip-direction of N20), that probably represent the main pathways for the water-circulation from infiltration zone to spring outlet. Structural and statistical analysis of outcrops analysed shown different geometric properties of Lower Jurassic Calcare Massiccio and Lower Cretaceous Maiolica Formations. Discrete Fracture Network (DFN) models of representative geocellar volumes have been built to compute both fracture porosity and correspondent permeability tensors (Kxx, Kyy, Kzz). Finally, the calculated fracture porosity for the Calcare Massiccio Formation is greater than in Maiolica Formation (1.7 and 4.3%, respectively), and the permeability values are well correlated with this trend.
Daily changes of dissolved inorganic carbon in river mainly are controlled by carbonate back-precipitation, aquatic photosynthesis and CO2 degassing at water-air interface. Dissolved inorganic carbon uptake by aquatic photosynthesis is a part of karst carbon sink, the size of degassing proportion values acts as the determinant factor for carbon sink stability. To analysis of the loss of bicarbonate caused by aquatic photosynthesis and relevant calcite precipitation processes, high resolution monitoring for water chemistry and high-frequency water sampling in daily scale were conducted in a 15 km-long section between Shengli and Guanyan in middle course of Lijiang River. The results showed that the flux of photosynthesis uptake of dissolved inorganic carbon in monitoring section is about 859 kgC/d. DIC removal value by photosynthesis and calcite precipitation value are 2.06 t/ d.km and 0.78 t/d.km respectively. CO2 degassing value account for 28.4% of total carbon removal, namely, approximately 70% of carbon removal was converted into organic carbon and precipitated in riverbed in form of calcite, thus constitute a part of karst carbon sink. DIC removal accounts for about 6.0 % of total input, in which 1.7 % returned to the atmosphere in the form of CO2, indicating DIC uptake by aquatic photosynthesis during the summer low water level can restrain CO2 degassing process in air-water interface effectively during the day time. The low degassing proportion suggested that DIC in water body of Lijiang River is relatively stable.
Mohamed GHANMIª,Amal MHIMDIª, Ines EZZINEª, Fadoua HAMZAOUIᵇ,Rachida BOUHLILAᶜ
ªLaboratory of Geoscience, Resource mineral, Energy and Environmental (LGREE), Geology Department, Faculty of Sciences of Tunis, University of Tunis El Manar, Tunis 1060, Tunisia.
E-mail : ghanmi.mohamed@gmail.com
ᵇLaboratory of Sedimentary Basins and Petroleum Geology (SBPG), LR18 ES07, Geology Department, Faculty of Sciences of Tunis, University of Tunis El Manar, Tunis 1060, Tunisia.
E-mail : fadoua_fst@yahoo.fr
ᶜDepartment of Civil Engineering, Modelling in Hydraulic and Environment, Laboratory, National Engineers School of Tunis, University of Tunis El Manar, Tunis, Tunisia.
E-mail : rachida.bouhlila@enit.utm.tn
Abstract:
Karst aquifers belong to the family of fractured aquifers. Within these reservoirs, complex flow processes exist as a direct consequence of the permeability contrasts affecting these heterogeneous environements. The objective of this study is to identify and characterize the geological and hydrogeological characteristics of the perched aquifers and carbonate, fractured and karstic reservoirs of the Jurassic located in the Zaghouan massif (NE Tunisia). The geological study shows that the Jurassic carbonates of Jebel Zaghouan form a NW dipping monocline. The inclination of the series is 30° and may still provide evidence of a ramped thrust at the base. Indeed, a major NE-SW fault collapses the NW compartment and contacts the Jurassic series with those of the Lower Cretaceous. This collapsed compartment extends to the southern border of the Jebel Mecella to the north (Jebel Rssas, Jebel Boukornin), the Zaghouan plain is Neogene and Quaternary filled. The Jurassic limestones of the eastern flank of the Jebel Zaghouan are bordered to the SE by the Zaghouan accident or overlap, locally marked by a Triassic lens. This fault places the Jurassic limestones and the marls of the Upper Eocene in contact. The Jurassic series of the Jebel Zaghouan are affected by intense fracturing with main faults of direction NE - SW, N - S, NW- SE and E - W. The fracturing has been observed at all scales and has facilitated the circulation of meteoric waters which have generated by dissolution a karstification system on the eastern flank of the Jebel Zaghouan.
The complexity of the Jurassic massifs of Zaghouan suggested a new tectonic system in this region, it is a tectonic style with accentuated overlapping at the level of Jebel Zaghouan and intense fracturing of the carbonates at the front of the overlapping allowing the development of karsts and the rapid circulation of underground water.
Key words: fractured aquifers, karsts, Jurassic massif, NW fault, overlap.
The study area is a part of the Saharan Atlas. Maastrichtian limestones constitute the aquifer through which manifest springs of fresh water. The limestone ridge separating the two major catchments with opposite flow (Chott Melghir catchment and Medjerda catchment) that emerge a number of springs.
The hydrogeological study relates to the relationship and the function of three springs located on different heights and spaced quite long distances. In the three springs, the variation in rates over time and the water level shows a resemblance, characterized by curves commonly affected by peaks which record maximum values.
The three springs studied have the same facies which is chlorinated to magnesium sulfate.
High concentrations of chemical elements are recorded during periods of flooding which shows the piston effect that characterizes the behavior of this environment.
The isotopic analysis shows a similarity with the meteoric waters reflect their short residence time and a lowest evaporation phenomenon of infiltrated groundwater.
At the edge of the Atlantic Ocean, one of the main complex karstic reservoirs of Morocco is located; it is the Cenomano-Turonian aquifer, which is drained by several sources along the Oued Iguerounzar. It is a source of drinking water supply and irrigation hence its socio-economic importance. Managing the water resources of this aquifer will help to mitigate the impact of overexploitation and quality degradation. Based on the results of the sampling campaigns of 1995, 2007, 2016, 2019 and 2020, an increase in mineralization is observed from 1995 to 2020. This mineralization comes from the dissolution of halite, dolomite, gypsum and sulphates but also from the effect of marine intrusion, which migrates towards the interior of the continent more and more in correlation with the intensive exploitation of groundwater. The flow direction is from SE to NW which explains the mineralization gradient which increases in the same direction. Two chemical facies are observed; the most dominant Cl-Ca-Mg and some points upstream show the SO4-Ca facies. The δ18O and δ2H isotopic values of the Cenomano-Turonian waters are respectively between -6.31 and 0.58 ‰ and between -39.2 and 2.2 ‰. The correlation diagram of oxygen-18 vs deuterium shows that the majority of the points are close to the GMWL and the LMWL, which translates a recharge by precipitations of Atlantic origin without significant evaporation between 740 and 1400 meters, which corresponds to the Kourimat region. Ultimately, this study reveals the interest of using isotopy and hydrochemistry to assess water resources and build solid knowledge in order to provide the keys for better management.
The development of urban zones in coastal areas leads to an increase in the demand for water resources. In this context, coastal aquifers represent an important source of water for supply in these areas. This is the case of the Yucatan peninsula (Mexico), made up of extremely porous and permeable calcareous sedimentary rocks, which favor the infiltration of rainwater, which is drained through submarine discharge. This coastal karst aquifer is the only source of drinking water supply in the Peninsula despite its high rainfall (1,100 mm/y). The Yucatan Peninsula is considered a regional-scale aquifer. Closest to the coastline, the aquifer is subject to the effect of the tides and a series of large sinkholes locally known as cenotes allows the connection between the surface and the groundwater. The main objective of this preliminary work is to analyze the hydrodynamic behavior of the freshwater in six of these cenotes, located in the coastal region of Playa del Carmen. All these cenotes have an irregular geometry, with areas that vary between 50 and 241 m2 and depth that reaches 4.5 m. The farthest is located at a distance of 700 m from the coastline.
The influence of the tide on the aquifer has been evaluated by continuous monitoring of the water level, temperature and electrical conductivity (EC) in two of the cenotes. Both cenotes are connected to the coast through an artificial channel made using a horizontal karst conduit that connected the cenote to the sea. This channel has also been longitudinally monitored. On the other hand, in-depth EC profiles were made in the rest of the cenotes studied. The EC varies between around 3,000 µS/cm in the most superficial strip of the cenote furthest from the coast and almost 20,000 µS/m in the deepest part of the one closest to it, with rapid increase with the depth in all of them. An EC-distance to the sea relationship is not observed, which indicates that the system presents some hydrodynamic complexity, probably as a consequence of the existence of freshwater discharges in some of these cenotes. Despite the proximity to the sea and the high transmissivity of the aquifer, a delay of up to 4 hours has been recognized in the response of the water level to the tide measured in Cozumel Island. The water level and salinity follow a correlative trend with a certain retard. Nevertheless, water level and temperature can show an inverse trend in some cases, as a consequence of the relation between the cenote discharge and the entry of water from the coastal area.
Abstract
In this study, we present the results of the hydraulic behavior of the Maastrichtian karst system of the Dyr perched syncline, that is situated in the Northern Est of the Algerian. It constitutes the North horst of the Tebessa subsidence basin. Processing and analyzing the hydro-pluviometric chronicles registered for 3134 days, between September 01st, 2002 and March 31st, 2011 at the hydropluviometric station of Ain Zerga, by means of the classified flow rates curves analysis method,
This curve is obtained by decomposing the hydrograph of the cycle into flow classes, and carrying for a given flow class, the number of events during which this flow was observed. The classified flow rates curves analysis aims to: highlight the phenomena that can affect the water flow of karstic system according to classification of (Mangin,1984) : “water leaks to the outside of the system, additional water supply to the system. This results in discontinuities. The distribution of these flow classes expressed as a cumulative percentage, is plotted on the ordinate on a probability scale, and the classes on the abscissa on an arithmetic or logarithmic scale.
As a result, we can admit that the functioning of thiskarstic aquifer system at annual time step is complex. This is certainly related to the complexity of the structure of this karstic system and the temporal irregularity of rainfall in this semi-arid regionKeywords: Karst spring, Ain Zerga, Tebessa, semi-arid, Flow rates curves
Keywords: Karstic spring, classified flow rates Maastrichtian,Ain zerga,Djebel Dyr,Algeria
Six tracer experiments were undertaken under different flow periods to delineate the catchment area and identify transport parameters in two snow-governed springs Laban and Assal in Mount Lebanon used for water supply. The two springs yield different responses to snow melt, ambient temperature in high flow and in recession despite their common origin from the same Albian-Cenomanian rock sequence. These discrepancies were attributed partly to different facies within the aquifer (limestone and dolostones). Yet faults and secondary fractures also play an important role in defining preferential flows in such a complex system. Secondary faults and fractures are difficult to depict in the field and were assessed via fracture analysis. In this work, primary faults with their characteristics (displacement and trends) are input in a Havana software (developed by Norsk Regnesentral; SAND 2021) based on field data used to simulate new faults. The model generates a secondary set of faults from a truncated fractal distribution, yielding thus different realizations of the set of secondary faults depending on the parametrization of the fractal model. The realizations will be validated with field data, doline distribution, and fracture analysis as well as tracer experiments results. This work allows to combine physical data with geostatistical techniques to optimize the delineation of the catchment and preferential flow in complex vulnerable karst systems.
In recent years, studies involving assessment of aquifer protection have advanced considerably. In general, an aquifer´s vulnerability depends essentially on the susceptibility of the vadose zone to offer greater or lesser protection from contamination. It also depends on the ease with which a contaminant applied to the land surface can reach the groundwater due to anthropogenic activities, environmental conditions and contaminant properties. Intrinsic vulnerability does not consider a specific contaminant and considers the susceptibility of groundwater to be contaminated by human activities and hydrogeological characteristics of an area of study. In tropical environments the higher rate of weathering allows the formation of thicker soils, which chemically and mineralogically differ drastically from those of temperate regions.The lack of studies about intrinsic vulnerability of karst aquifers in South America and Caribbean can be a problem in many countries, so this research investigates various methodologies and the applicability of intrinsic vulnerability of karst aquifers developed on carbonate rocks. South America and the Caribbean have almost 5% of the world´s carbonate rock. However, some countries have large extensions of their territory covered by karst rock, such as Cuba, 67%, and Mexico with 25%. It is estimated that more than 10 million people use water from the karst system in Mexico. In Cuba, 33% of all available water volume originates from groundwater, and 92% originates from karst aquifers. Brazil, Peru, and Argentina also appear in that order of importance regarding carbonate rock water resources. Other countries that in the project have reduced importance in terms of area and fewer studies of intrinsic vulnerability to contamination. The countries with the most studies about the protection of karst aquifers are Mexico, Cuba and Brazil. Studies in Mexico are mostly concentrated in the Yucatán, where a specific methodology has been developed that bears the name of the region-IVAKY. In Cuba, Brazil, and Colombia, the methodologies applied were based on the European approach known as COST Action 620. In the other countries in the study, exclusive methodologies for protecting karst aquifers are mostly unknown.
Keywords: karst aquifers, intrinsic vulnerability, carbonate rock, Caribbean, South America.
The effects of the main climate variables, rainfall and temperature, on the groundwater conditions have been investigated by different statistical analyses on long hydrological time series. Data refer to the main karst springs of Campania, southern Italy, and local climate stations.
The first approach has been the standardization of the series; this procedure requires the fit of the observed data by different three-parameter probability distribution; then, an accurate statistical criterion has been used to select the most suitable probability model to calculate robust hydrological standardized series, in order to examine duration and intensity of hydrologically dry and wet periods. Hydrological droughts have been focused, considering their role on the economies and environment.
After the distribution fit has been found, the time series has been transformed into the standard normal distribution; for the rainfall time series this provide the well-known SPI and for the discharge time series a new index, the Standardized Discharge Index, has been proposed.
The second approach has been the estimation of trends by least squares linear regression and the Mann-Kendall and Sen’s slope tests.
A strong relationship between climate variations and groundwater system has been found; in particular, the intensity and the frequency of hydrological droughts have been increased over the last three decades, and a statistically significant decrease of the annual mean spring discharge was observed, together with a strong statistically significant trend of the annual mean air temperature. The latter had been a strong impact on the karst aquifers and induced an increase of duration and intensity of hydrological droughts in the last decades. On the other hand, no statistically significant long-term trend seems to characterize the annual mean precipitation series.
In Mexico, the main source of water supply for human use is found in subway areas, a large part is in the aquifer of the Yucatan Peninsula, however, there are already studies that report contamination by different pollutants, so it is necessary to know, analyze and evaluate the threats, vulnerability and risk to contamination to propose strategies for the conservation of the most vulnerable areas. Therefore, the objective of this research was to develop an index of vulnerability to contamination of the coastal karst aquifer for the urban zone of the city of Playa del Carmen, utilizing remote sensing, considering 4 factors: A) The relief, measured with three indicators: the density of karst depressions, depth of the depressions and the type of depression (sinkhole, uvala or polje) from the processing of LiDAR data of 5m resolution in ASCII/GRID format of the INEGI. B) Climate, measured by rainfall intensity, with the Modified Fournier Index (MFI), from the interpolation of the average of 28 years (1990-2018) of precipitation from 11 meteorological stations. C) Land use, with the secondary zoning of the Urban Development Program of the Population Center of Playa del Carmen. And D) Sea level rise, with three indicators: Estimated projections of 1, 3, and 5 meters above sea level, distance to the coastline, and piezometric level. The results showed a higher degree of vulnerability in the first three kilometres, with a tendency towards the centre of the urban area. This zone is characterized by an average of 17 karst units per km2, with a predominance of sinkholes, with a depth greater than 4.8 m and located at a distance of less than 3 km from the coastline; this zone is intended for residential, tourist, residential and commercial use and is classified within the zone of very high precipitation according to the classification proposed in the MFI. The limitation of this study was the scarcity of data on precipitation and piezometric levels that would allow a more detailed analysis; however, the remaining indicators have a high degree of reliability.
Red Lake near the town of Imotski in Croatia is one of the most impressive examples of karst features in the Dinaric Karst. This area is rich in karst formations, and Red Lake is one of the least explored. The reason for this is the difficult access to the lake, which requires more complex, demanding, and, above all, resource-intensive research. Therefore, the analysis of recession processes in the lake allows quantification of hydrological phenomena and insight into the hydrogeological functioning of the lake. The analysis is based on data on precipitation and water levels in the lake. Integral water inflow and outflow quantities were determined using the existing morphometric model of the lake, which defines the relationship between lake volume and depth. The average values of the recession coefficients, defining the dominant regime of the lake, were obtained from Master recession curves. The Master recession curve provides information on the average characteristics of the discharge components and was determined using 3 methods: Adapted Matching Strip Method, Tabulation Method, and Petraš Method. Since the object of study in this paper is a lake and not a watercourse, the strict definition of recession curves can not be applied. The recession periods of hourly water volume change of the lake were isolated, normalized, and log-transformed, therefore the analyzed periods are referred to as quasi-recession curves and the recession coefficients as quasi-recessions coefficients. Analysis of quasi-recession curves and calculated coefficients, define the relationship between the direct and baseflow component, further on they identify the predominant hydrogeologic mechanism as well as the response of the lake to the precipitation and the lake's connection to a nearby spring. This approach can be used not only to identify the hydrogeological regime of karst lakes but also to stratify the different porosity levels of the surrounding karst massif.
The shores of the Mediterranean basin are marked by the presence of karstic watersheds subject to exceptionally intense rainfall events. Modeling the hydrodynamic functioning of these watersheds at a rainy event scale is crucial for determining warning thresholds and sizing engineering structures to protect people and property (Bonacci et al, 2006). The role of karst in flood dynamics is often neglected because it remains difficult to estimate: the processes are highly non-linear and data acquisition is very delicate. Thus, the volumes involved remain difficult to assess, particularly because of the spatial variability of precipitation, the high uncertainties in flow measurements and the heterogeneity of flows (Bailly-Comte et al., 2008). Also, works on this topic have generally favored an integrative approach based on systemic or conceptual approaches integrating or not a karst component (Coustau et al., 2012; Kong-A-Siou et al., 2013; Raynaud et al., 2015) which may be insufficient depending on the characteristics of the basins.
By its geological and geomorphological characteristics, the watershed of St Pierre-la- Vacquerie located on the southern edge of the Larzac plateau offers the possibility to access the dynamics of the role of karst. Indeed, with an area of about 1 km2, it is mainly composed of dolomites and limestones with little karstification except in its downstream part where two avens are located on faults. It is equipped with a meteorological station and its outlet corresponds to the main street of the village of La Vacquerie-et-St-Martin-de-Castries which becomes a reach where the flow can be measured during floods.
The modeling consisted in simulating the hydrological behavior of the basin during intense rainy episodes with the HEC-RAS software. For the topography of the basin, a metric DTM was used. Infiltration coefficients of the different soil types were measured and then spatialized from land use maps and drone aerial views. Measurements of an intense rainfall event in 2015 were integrated into the model to simulate the flooding of the main talweg whose characteristics were mapped. After passing through two channels, the stream is channeled through the main street whose dimensions and Strickler coefficients were determined. The flows in the street could be estimated from the water heights and from the analysis of videos taken during this episode (FUDDA LSPIV software). These data were used to calibrate and validate the model. It was thus possible to estimate the volumes absorbed by the sinkholes at each time step and thus the dynamic participation of the karst. In terms of crisis management, a threshold of rainfall intensity was proposed at which the karst is saturated and the village flooded.
The exploitation of underground water has an important role in the province of Alicante, Southeast Spain, where it’s mostly used for agriculture and the supply of populations. For example, the Almudaina-Segaria hydrogeological domain, between the Marina Alta and El Comtat districts, in northern Alicante, constitutes the water supply for the La Vall d’Ebo, Planes y Gorga villages.
This domain includes 9 aquifers, with a total surface of 215 km2. 145 km2 correspond to the Almudaina-Segaria system, a carbonated karstic aquifer with 700 m of permeable thickness made of limestones, dolomites, marls and calcarenites. The system has no hydraulic connection with the surrounding aquifers due to the presence of impervious facies, except in the far Eastern sector where it is linked to the detrital materials from the Vergel and Pego aquifers.
The water input is provided by the infiltration of precipitation (approximately 730 mm and 65 hm3 annual precipitation and recharge, respectively), while the output occurs through direct pumping (up to 7 hm3/yr), natural drainage to the Vergel and Pego aquifers and, mainly, discharge through the Balsa de Sineu spring and the Racons river (58 hm3/yr).
In this work we modelled the aquifer’s water flow dynamics aiming to understand its current state and how it works. The building of the numerical model was carried out using the MODFLOW code and the ModelMuse graphical user interface. The model was designed considering the pre-existing models, updating the information until 2019. We obtained then an updated model and its estimated evolution for the period 1980-2019.
The groundwater flow occurs in a SW-NE direction, in accordance with regional tectonics, towards the discharge zone. The results show an important temporal variation in the piezometric levels in the western portion of the aquifer, with variations over 50 m throughout the evaluated period, in accordance with periods of greater and lower rainfall. In this sector, the piezometric surface is between 350 and 250 m.a.s.l., and it descends in the flow direction reaching values even below the sea level in the far east. In this last sector, the level remains relatively stable through time.
For the model calibration, the results were compared with the measurements of the available piezometers that represent different sectors of the aquifer, throughout the entire time series. The root mean square error (RMSE) is roughly 5%, and every model step has a discrepancy percentage smaller than 0.05%.
The modelling of the aquifer is a useful tool for an appropriate management of water resources in the north of the Alicante province, given its importance as a main support for the populations and activities that take place there.
This study deals with the vulnerability and pollution risk in the East of Tebessa plain aquifer (Bekkaria), which situated in the extreme Algerian East, where a semi arid climate reigns.
For specifying the degree of vulnerability of our study area, we have applied the mapping of tow methods: DRASTIC and GOD; which allowed us to determine low, midium and high areas vulnerability.
The obtained vulnerability map made with the DRASTIC method shows three zones of different vulnerability degrees: areas of low and medium vulnerability occupy the majority of the total area of the plain, while high vulnerability areas occupy a small superficy in some places. In the case of method GOD, the areas of low vulnerability dominate the map, it has also a medium vulnerability towards the center of the study area and finally a high vulnerability near the town of Bekkaria.
Keywords: Vulnerability, Pollution, Tébessa, Bekkaria, DRASTIC, GOD,
The north-western area of the Classical Karst (NE Italy), is subject to frequent flooding after abundant precipitations, which become problematic since 2001 as they more frequently affect housing and recreational areas, leading the population to believe that the swallow holes draining the area stopped functioning. The climate changes as well as the increased frequency of intense rainfall events led to study more in depth the situation in the Mucille karstic depression. The latter is fed by a spring area and drained by two swallow holes one of which is permanently active while the other functions only during floods. About 24 hours after the onset of heavy rains, the whole depressed area is flooded. About 8 days later, the water level begins to decrease, coming back to its initial height in about fifteen days. During floods, while springs and swallow holes discharges measurements are impossible, the extension of the flooded areas has been mapped. The obtained flooded surface together with high resolution DEM coverage allows to calculate the volume of surface water. Combined with water table levels registered in an adjacent piezometer, this volume can be computed over time. Consequently, the hydrologic balance (inflow minus outflow) can be estimated during the whole event. The first day of the flood that occurred during December 2017 (100 mm of rain in less than 48 hours), the difference between the inflow and the outflow averaged about 900 l/s. The following day it decreased to 280 l/s. The next 5 days the water balance was close to equilibrium. From the eighth day on, outflow became predominant resulting in a negative budget between -240 and -80 l/s.
This study provides evidences fundamental for the design of measures to mitigate the risk. It estimates the discharge of the swallow holes, confirming their efficiency. Nonetheless it also emphasises the need to improve their draining capacity, especially considering the unsuspected high outflow of the springs at the onset of the flood.
Climate change and pollution are posing additional unprecedented threats to existing water resources, especially to water supply from karst aquifers in Mediterranean and semi-arid regions. A numerical model considering the most important key hydraulic parameters can forecast the impact of any given input on model quality and quantity output. In this work, we propose to model flow and transport using Comsol multiphysics in a synthetic model and to apply it to a simplified real case study (Jeita spring in Lebanon supplying water to 1.5 million inhabitants). The model geometry consists of a 5300 m long variably saturated horizontal conduit portrayed as 1) 2-D continuum and/or 2) a channel draining a porous equivalent matrix (400 m thick). Flow is simulated using the Richards Equation in both saturated and unsaturated medium. Recharge is applied vertically as both diffuse and point source in a shaft linked to the conduit. Percentages of fast infiltration rates are obtained from the analysis of event time series recorded at the spring (electrical conductivity and discharge). Flow rates at the outlet are used for transient model calibration. Mean velocities, dispersivities, and phreatic conduit diameters obtained from tracer experiments under various flow periods are used for transport validation in the channel. The aim is to test the validity of a functional simplified flow model on a complex real case and to identify based on a sensitivity analysis the key parameters that allow an optimal calibration of such a model.
The karstic Trinity Aquifers of central Texas provide baseflow to streams and are used extensively as water supplies for domestic, municipal, agricultural, and industrial purposes. Rapid increases in population in the area are placing significant demands on the aquifers in an area that has very limited surface water supplies.
The goals of groundwater management agencies in central Texas are to maintain adequate pressure and water levels in the aquifers, protect the yields from water-supply wells for users that are 100% dependent on groundwater, and to maintain sufficient flow from springs to protect the ecology that is dependent on this flow. Studies have indicated that decreased streamflow and water levels in the Trinity Aquifers can reduce recharge to the Edwards Aquifer that is downgradient to the Trinity Aquifers. Some of the tools being used to characterize these aquifers are a network of transducers in monitor wells, synoptic water-level measurements, streamflow measurements for gain/loss studies, geochemical analyses of groundwater and surface water, aquifer testing, dye-trace studies, and analytical and numerical modeling.
Droughts occur frequently in this area and current levels of pumping in the Trinity Aquifer have resulted in both the capture of springflow in karstic areas, resulting in a major spring ceasing flow during drought, and groundwater mining in other less karstic areas of the Trinity. The most significant historical drought in central Texas occurred in the 1950s and lasted for up to 10 years. Tree-ring data show that even more significant droughts have occurred over the past thousand years. Predictions for changes to precipitation due to climate change are for more extensive flooding and more severe drought. Long-term trends in water levels are downward with only limited recovery during very wet periods.
Water discharging from Trinity springs provides for recharge back to the Trinity through downstream karst features in the stream beds. Flow through conduits has been documented in some parts of these aquifers, but little is known about the distribution of flow through conduits versus flow through the rock matrix and flow through a significant fracture network. The certain increase in demand for groundwater coupled with severe drought will likely cause significant further reduction in springflow and decreases in water levels. This will cause many water-supply wells to cease the ability to yield water and negative ecological impacts along spring-fed streams, which will both lead to serious economic consequences for the area. Furthermore, reduction of Trinity Aquifer springflow will reduce recharge to the Edwards Aquifer that hosts a variety of endangered aquatic species at Barton and San Marcos Springs.
An understanding of how these aquifer systems function is key to proper aquifer management which is likely to involve limits to pumping during average conditions and more severe reductions during periods of drought, and incorporating alternative water supplies such as rainwater harvesting, aquifer storage and recovery, and perhaps importation of water from distant sources.
Karst aquifers of southern Italy are huge reservoir of groundwater, with variations of storage during the hydrological year, from a minimum amount in summer-autumn season, and maximum amount during the spring season.
In recent years it has been observed that these hydrological cycles are associated to observable displacements (mostly in the horizontal component) measured by GPS observations on ground surface, that modulates the active tectonic extension across the Apennines (~3 mm/yr). Besides, these different hydrological stages, show a robust correlation with rate of seismicity as measured by local seismic networks.
Based on continuous GPS measurements since 2005 around the Matese karst massif in southern Italy, a strong correlation between horizontal displacements and karst spring discharge has been found; these results highlight the important role of the hydraulic head into the saturated zone of the aquifer in modulation of the stress and the strain accumulation in the seismogenic and tectonically-active upper part of the Earth crust. As consequence of this behavior, also the rainfall accumulated over specific time interval has the same role.
It has been also found the high correlation between the high flow conditions of springs and the rate of low-magnitude instrumental seismicity, which highlights the important role of the hydrological conditions on triggered seismicity.
Big Spring, located in the Ozarks region (Carter County, Missouri), is one of the largest karstic springs in the United States and the world in terms of flow. With an average flow of 12 m3/s and an average daily discharge of 1.106 m3, this spring is a first magnitude spring which rises at the base of a dolomite cliff at the west side of the Current Valley, Ozarks.
To assess long-term projections of climate change and short-term evolution of anthropogenic impact such as pumping on Big Spring discharge, we analyzed the daily discharge records collected by the USGS from 1921 through 1996. The first step of the assessment was to use spectral and correlation analyses to analyze the daily discharge data. A correlogram was calculated using the daily discharge records, which showed a memory effect of 95 days, indicating the importance of storage capacity of the Big Spring karstic aquifer. A cross correlogram between rainfall and discharge data was conducted and showed a differentiated effect with delay between rainfall and snow.
The second step consisted in deseasonalizing the discharge data using a centered moving average of three, six and twelve months. The resulting data was analyzed with spectral analysis and wavelets techniques to assess the long-term effect of climate change on Big Spring’s storage capacity.
Karstic aquifers in Andalusia are one of the most important freshwater sources for the region containing more than 2000 hm3/year of renewable water resources. Although these aquifer systems have been widely studied, a regional systematic study has not been carried out to explain the spatial variability of the hydrochemical characteristics of groundwater and its relationships with climatic, geologic and biotic variables.
In this study, a regional assessment of all karstic aquifers was carried out, selecting 40 springs of the most representative in which the waters were sampled. In addition to the in situ pH, temperature, and electrical conductivity determination, the hydrochemical and isotopic composition of the waters and the isotopic content of 13C in DIC have been determined. At the same time, recharge area, recharge rate, vegetation cover (Normalized Difference Vegetation Index, NDVI) and geological characteristics of the system, have been determined for each aquifer. Except in the thermal springs, the waters mainly have temperatures and electrical conductivities between 10 and 20 oC and 300 and 750 µS/cm, respectively. The most common hydrochemical types are HCO3 to HCO3-SO4 and Ca to Ca-Mg.
Finally, processes governing water mineralization were analyzed using multivariate statistics (Principal component analysis). Although the hydrochemical composition of the groundwater is relatively homogeneous, differences are found mainly related to the geological domain in which the springs are located. Moreover, from preliminary results, we obtain that the hydrochemical variability is also influenced by environmental factors related to the water temperature, the recharge rate and the vegetation cover. We observe a good correlation between the air temperature, the water temperature, the recharge rate and the isotopic composition of the water and the DIC. Another important variability factor is related to the NDVI index, the apparent equilibrium partial pressure of CO2 and the bicarbonate content.
The Cevennes national Park (PnC) and the French Geological survey (BRGM) have worked together to study the karst groundwater resources of the Causse Méjean, France. This karstic plateau is made up of Jurassic carbonate sediments (Lias, Dogger, Malm). ). Except the liasic marls (Toarcian and the Domerian), it is made of limestones and dolomites that gives a typical karst landscape all over the Causse and its adjacent valleys.
The Causse Méjean has undergone a long period of weathering under cover, producing deep corridors that are mostly still filled with alterites (dolomitic sands). In the western part of the Causse , discharge monitoring shows that this ghost rocks karstification controls the highly inertial hydrodynamic behavior of the phreatic zone. However, the current karstic functioning exports these alterites, which allows the rapid development of a limited number but well-developed karstic networks. This explains the predominance of convective transport revealed by tracer tests within these pseudo-endokarsts. To the East, the karstification was enhanced thanks to allochthonous recharge from surface stream draining the Cevennes mountains. Associated with the dip of the layers towards the West, this configuration explains the paleo-direction of the karst drainage revealed by tracer tests that reach the Tarn River. This major direction of flows was then modified due to the deepening of the Tarn, the Jonte and finally the Tarnon rivers. The deepening of the Tarn River allows the emergence of karst springs that reorganize the karst drainage to the North (i.e. Castelbouc Sp). The deepening of the Jonte River explains the relatively limited karstic drainage of the Causse Méjean towards the Douzes sp., which is mostly recharged by swallow holes in the Jonte riverbed. Finally, the deepening of the Tarnon river through all the Jurassic sedimentary sequence explains the emergence of the Pêcher Sp. in contact with the substratum. Results of 22 tracer tests conducted during the project validate this karstogenetic evolution. In addition, the deepening of the Tarnon river has also hydraulically disconnected fluviokarst streams and polje systems from their Cevennes watershed. Materials of Cevennes origin still remain in karst depressions and filled the vadose and phreatic zone of the karst systems. Geochemical water-rock interactions with these materials explain radiogenic anomalies of Sr isotopes. This interpretation has been used to better understand the origin of the water flowing to the Pêcher spring, and to assess the relative contribution of the Jonte River to the discharge of the Douze Sp.
This study shows how knowledge of the geological and geomorphological evolution of the Causse Mejean can be used to build a karstogenetic model to be challenged by other hydrogeological tools and methods. This end up with a consistent representation of the structure and the hydrological functioning of the karst systems at the scale of the Causse Méjean.
Karst aquifers are exposed in many regions of the Mediterranean and represent important freshwater resources for drinking water supply and irrigation. Their accessibility varies by geographic location and is coming under increasing pressure from population growth and climate change. Sustainable management of these unique aquifers at local, transboundary, and transregional scales is therefore of great importance and requires a consistent representation of geologic, hydrogeologic, and water management related information. This study demonstrates the development of a consistent Mediterranean karst aquifer map and database (MEDKAM) consisting of multiple maps, layers, and associated geospatial information. MEDKAM is part of the WHYMAP series under the umbrella of UNESCO and BGR, which ensures compliance with international map quality standards and wide international dissemination. The principal karst aquifer map is generated by a reclassification of the international hydrogeological map of Europe (IHME) and the recently re-digitized parts of North Africa and the Arabian Peninsula. The map was updated to include new regional karst maps and contains more detailed information on the spatial distribution of various karstifiable rocks, major karst springs, caves, and karst-groundwater-dependent ecosystems (KGDE). MEDKAM also contains basic information on other aquifer types. Additional advanced analyses were conducted as part of MEDKAM, such as on groundwater stress and pollution risk. The groundwater pollution risk was addressed by converting global land use data into categories of low to high risk for groundwater pollution, while groundwater stress was assessed by analyzing trends in changes of groundwater storage using the latest data from the Gravity Recovery and Climate Experiment (GRACE) satellite mission and newly analyzed ERA5 land climate data from the European Centre for Medium-Range Weather Forecasts (ECMWF). The maps and database allow users to perform a variety of analyses at regional and supra-regional scales, and also to place local studies in a regional context. In addition, MEDKAM will serve scientists and decision makers as a basis for transboundary decision making. The principal map will be available in printed form and online for download.
Groundwater tracer tests to five public water supply abstractions in the Upper Cretaceous Chalk of South East England are reported. The tracer tests are in an area of well-developed karst in an environmentally-stressed, peri-urban region north of London. The tests were conducted from three geographically distinct areas of surface karst development as part of an investigation into the influence of rapid groundwater flow to public water supply abstractions. Rapid surface-to-source contaminant transport in karst is a significant risk to water quality as the impact of natural attenuation is severely limited. Characterising the influence of rapid groundwater flow at public water supply abstractions in areas of well-developed chalk karst therefore has important implications for catchment management strategies and the delineation of groundwater source protection zones.
There is a requirement to employ the precautionary principle when tracer testing to public water supply abstractions. Accordingly, a new methodology is reported to allow for safe dye tracing to public water supply abstractions. The suitability of bacteriophage tracers in this setting is also appraised. This presentation reports: i) the results of recent field testing using dye and bacteriophage tracers; and ii) the implications for groundwater source protection zone development in the field area and the wider Chalk aquifer in England. It also evaluates the new testing methodology.
Springs are important to the humanity because they provide potable water to many locations in the world, thus ensuring health, sanitary conditions, food production and economic development. Karst and mineral water springs are particularly important, but springs emerging from karst aquifers are by far the largest – some are even discharging entire underground rivers, whose flow sometimes exceeds 100m3/s. Having caused the establishment of permanent settlements and nearby cities, especially in the Roman times, many springs are also historically important.
Although some of the best known karst springs are actively used and very well protected from pollution, many others around the world have been contaminated, devastated by over-pumping or impounded by reservoirs. Taking advantage of the opportunity provided by the 50th anniversary of the IAH Karst Commission (KC) and the involvement of UNESCO in our karst research, this initiative (project) aims to bring together both the KC members and many national experts to work, on a voluntary basis, to: 1. develop criteria for the selection of most important karst springs, which inter alia should include historic, aesthetic and scientific values; 2. establish the list of springs; 3. create the Code of Practice for these springs’ utilisation and protection; and 4. promote these springs by their in situ labelling and internet publicising.
The idea to identify and protect selected springs does not imply prevention of their further use. To the contrary, this initiative intends to highlight their importance, defend them against possible devastation and ensure that any further intervention consider their protective status.
The Eurokarst 2022 and the KC annual meeting would be an excellent opportunity to present and discuss the project’s concept, create Advisory Board with regional representatives and trace further steps.
Karst aquifers constitute an important water resource and ensure the supply of drinking water to certain French metropolises such as Montpellier, Angoulême, Poitiers, Cahors but also many cities around the Mediterranean, such as Rome (Italy), Beirut (Lebanon) or Malaga (Spain). In the current context of climate change, groundwater resources could be significantly impacted, in particular by a modification of groundwater recharge processes. Water withdrawals resulting from human activities (domestic, industrial or agricultural uses in particular) also constitute a factor of increased anthropogenic pressure on the quantity of groundwater and raise certain sensitive questions for users and managers: how to quantify the impact of water withdrawals on the available water resource? How to guide managers in an adequate distribution of resources between the various users while preserving the reserve flows of rivers, especially during low water periods?
The development of innovative approaches in the field of hydrological modelling should make it possible to estimate the impact of human activities and attempt to predict potential future developments. In recent years, the rainfall - level - flow modelling platform KarstMod, developed within the SNO KARST (French national karst observation service, INSU-CNRS), has made it possible to meet the modelling needs of flow processes in karst aquifers. This tool allows to calibrate a hydrological model using all the observed data available (precipitation, evapotranspiration, flow at the outlet, pumping flow, piezometric level). The use of such a model allows i) to appreciate the internal dynamics of flows (exchange flow between the different compartments of the aquifer), ii) to simulate a hydrograph from precipitation and evapotranspiration time series included in the context of climate projections and/or scenarios of groundwater abstraction and iii) to estimate the parametric and predictive uncertainties associated with the simulations. This model and its functionalities thus constitute a relevant tool for evaluating the impact of i) groundwater withdrawals on the internal dynamics of the system and therefore the functioning of the aquifer, ii) changes in land use on recharge and iii) climate change on the availability of water resources. We propose a synthetic description of the approaches currently developed within the SNO KARST and review the main results obtained on some karst hydrosystems in metropolitan France: the Touvre (Charente), the Fontaine de Vaucluse (Vaucluse), the Lez (Hérault) and the source of l'Oeillal (Aude).
Groundwater in karst aquifers is a valuable resource in Mediterranean region which is particularly threatened by climate change. In fact, climate change induces an increase in evapotranspiration and a decrease in rainfall available for groundwater resource. This impacts directly karst aquifer recharge and thus indirectly karst-river interactions. The study site is a karstic catchment located in the Mediterranean french river basin and known as the “Karstic Gorges of the Cèze River”. This site has been studied through interdisciplinary research since 2014 to improve understanding on how the catchment is functioning on different aspects. In fact, the river at different points and each of its karstic springs have been equipped with probes and studied by multiple disciplines (geochemistry, biology, radioactivity…). As a continuation of these researches, this communication presents an attempt to model the interactions between the whole karstic catchment of the Gorges and the Cèze river and then to predict their evolution regarding to climate change scenarios. The model has been developed using Karstmod software, which consists in an adjustable modelling platform. Firstly, the model has been calibrated on one hydrological cycle and validated on the following cycle. Then future climate dataset simulated by the ALADIN model, according to Intergovernmental Panel on Climate Change (IPCC) Representative Concentration Pathway (RCP) scenario data, and corresponding to the study area were injected in the above-named model in order to predict karst and river interactions and water production for the period 2030 - 2100. The results of this work are presented in terms of volume of water produced for three simulations corresponding to three RCP scenarios (2.6, 4.5 and 8.5).
For karst aquifers, it can be assumed that sporadic groundwater sampling only provides an incomplete picture of pollutant occurrence and variations. For Switzerland, the frequency of maximum 4 samples per year in the framework of the long-term groundwater monitoring thus probably only insufficiently reflects the concentration dynamics of pesticides and other micropollutants in karst areas. In contrast to unconsolidated aquifers, also substances that are relatively easily degradable and less mobile can be detected at karst monitoring sites. The present pilot study, which is part of the Swiss Action Plan for Risk Reduction and Sustainable Use of Plant Protection Products, therefore aims at assessing the short-term concentration variations of pesticide concentrations in karst groundwater.
In a first step, in order to identify suitable springs that have the potential for showing short-term pollutant peaks, 14 days composite samples were obtained at 10 karst springs in the Swiss Jura Mountains from March to October 2020. Subsequently, three of these springs were selected to be monitored with 42 hours composite samples from March to October 2021. Both sampling sets were screened for pesticides and related transformation products at the laboratory. In an additional step, at one spring so far, the occurrence of concentration peaks was investigated by means of a mobile mass spectrometer unit (MS2field) which allows for analyzing pesticides and their transformation products automatically and in-situ every 20 minutes. These were recorded over a several weeks period including distinct precipitation events, in conjunction with the continuous measurement of other parameters such as electrical conductivity, turbidity, water temperature nitrate, pH or bacterial cell count (online flow cytometry).
The high temporal resolution of the MS2field data allowed the detection of significant pesticide peaks directly following recharge events, which were less obvious during routine monitoring. In conjunction with a detailed assessment of land-use activities, including information on the application of pesticides in the catchment, the findings provide more insight into subsurface transport processes and explain the occurrence of pesticides at the related spring. Consequently, the study gives information on how sampling strategies may have to be adapted in order to improve the monitoring of micropollutants in dynamic karst systems.
Wetlands provide important regulating ecosystem services, in addition to providing rich habitats for biodiversity and yet many of the world's wetland ecosystems have been removed, whilst others are under threat due to proximal land degradation, water quality pollutant impacts and/or water supply pressures, making them among Europe’s most threatened ecosystems. Understanding the eco-hydrogeological connectivity and environmental supporting conditions is critical to the management of wetlands, as well as meeting legislative commitments, such as the implementation of the Water Framework Directive (WFD), particularly for those wetlands defined as groundwater-dependent terrestrial ecosystems (GWDTEs).
Four calcareous fens supported by groundwater from karst aquifers in Ireland were instrumented and monitored over a two-year period, investigating both their hydrology and hydrochemistry in relation to their different vegetation habitats. The fens were selected to exhibit a range of different water quantity (drainage) and water quality (nutrient) pressures. Regular water quality monitoring was carried out across different transects on the fens comparing samples taken at the free water surface (phreatic), from the water at depths (3 to 8 m) at the transition zone between the peat layer to underlying subsoils (piezometers) and from groundwater in the aquifers feeding the fens. The results of water balance calculations and the hydrochemistry monitoring provided a conceptual model of groundwater feeding the fens at discrete points, helping to maintain high water levels even across drought periods in the summer, as well as supplying relatively high concentrations of nutrients which appear to be picked up by the fen vegetation, thereby leaving lower nutrient concentrations in the surface water runoff from the wetlands. The vegetation dies back annually, accumulating to form peat as well as partially degrading, thereby performing a key role within the in-fen nutrient and organic cycles within the underlying substrate.
In terms of water level, the field investigations across all fen habitats (PF1, PF2 and PF3) suggest that a threshold water level envelope of between 30 mm to 275 mm above ground level need to be sustained for at least 60% of the year is required for healthy fen vegetation with the mean annual water level always above the surface. In terms of water quality, the setting of thresholds is more challenging and is an area that needs further research, particularly with respect to the link between the wider supporting catchment groundwater water quality and quantity metrics and those metrics defined in the wetlands for different vegetation types. Whilst envelopes of nutrients in the surface water of the fen associated with healthy fen vegetation can be defined (for example, dissolved reactive phosphorus concentrations of between 6 to 37 µg-P/l for PF1 habitat), these levels should be regarded as the water quality after the fen vegetation has effectively treated the higher incoming nutrient levels in the groundwater. At the four fens studied the generally higher levels of nutrients in the groundwater feeds did not appear to be causing any ecological stress to the fen ecosystems and so groundwater threshold values cannot yet be defined with any confidence.
Since the roman era, the karst springs of Zaghouan, supplied drinking water for the local cities and for the capital Tunis. Aware of its social, economic and cultural importance, it is undeniable for us to understand its hydrodynamic behaviour. The mountain contains 14 springs. The most important springs are located on the mountain north-western slope among them: Nymphea, Ain Ayed, Ain ElGuelb, Gallery 44 and Gallery 47 and Ain Haroun.
The adopted methodology is based on 1) processing and statistical analyzing of the variables related to the karst: rainfall, pressure, temperature and discharge 2) modelling the rainfall-discharge relationship with artificial neural networks (ANN).
Available Graphical discharge series at the spring of the Nymphea, dating from 1915 to 1944 at irregular time scales, were digitized. This digitalization allowed us to obtain a complete and continuous discharge series on a weekly scale. The discharge on a daily scale was obtained by linear interpolation. For temperature and pressure series, historical series are not available, we used a synthetic temperature and pressure series (median series) built from series from 2019 to 1944.
We considered three types of ANN models: Multi-layer perceptron (MLP), Long Short Term Memory (LSTM) and Convolutional Neural Network (CNN). In addition, in order to characterize the most suitable time scale for our system, we performed daily and weekly simulations for each model. For each temporal scale, a simulation configuration has been adopted: it consists in predicting a single future value of discharge knowing only the data on rainfall, average temperature and pressure of the previous days. This configuration is called "seq2val forecasting" or "one-step ahead forecasting". In order to prevent overfitting, we used the "dropout" regularization method which consists in randomly deactivating some neurons at each iteration. Our different models have hyperparameters which are the guiding part of the simulations. However, it is not obvious to find the optimal combination of hyperparameters. Therefore, we used Bayesian optimization to ensure optimal performance
Our modelling approach consisted of two phases: a calibration phase and an evaluation phase. Data was thus split as follows: 85% (24 years) and 15% (4 years) of the data was used in the calibration and the evaluation phases respectively.
The evaluation parameters of our models were: R², NSE, Bias, RMSE, rRMSE, rBias.
For all considered models, the simulations performed on weekly data provided better results than the simulations on daily data. We deduced that the weekly scale is the most relevant in the hydrodynamic study of the Zaghouan karst. This is due to the fact that (1) the daily discharges were obtained by interpolation of the weekly discharge, (2) the cumulative uncertainties of the discharge measurements and (3) the cumulative uncertainties of the daily rainfall measurements, some of which were taken every 3 days. Furthermore, the MLP model gave the best results on the weekly test set (R²=0.65; NSE=0.68) compared to the CNN (R²= 0.32; NSE=0.66) and LSTM (R²=0.17; NSE=0.72) models. LSTM gave the worst results. The distribution of predicted weekly discharge in relation to observed weekly discharge shows an overestimation in our forecasting. Notwithstanding the use of a synthetic temperature and pressure series, our results were on the whole acceptable. Besides, we looked for the lag time between the rainfall series and the discharge series (to find the reactivity duration) by maximizing Kendall's tau coefficient while checking the significance test. For Both daily and weekly scales, the lag time was about 18 weeks.
Abstract
The issue of water is now at the heart of the concerns of the countries bordering the Mediterranean, and the efficient management of water resources is becoming one of the major issues around the Mediterranean. These water resources are surprisingly reduced due to pressure from users, as well as climate change which is a factor aggravating water stress.
In Morrocco, the Middle Atlas region contains karst springs of great importance, with flows exceeding 2.5 m3 / s. In this region, groundwater of karst origin plays a major role in increasing the flow of Oued Sebou, considered the great river in Morocco. These sources experience remarkable seasonal fluctuations, where it dries up in periods of long droughts.The main objective of this study is to obtain a better knowledge of the hydraulic functioning of the karst complex which is located downstream of the Haut Sebou watershed.
Key Words : karst springs, hydraulic functioning, seasonal fluctuations, water stress.
Today, the impact of climate change on water resources has become a tangible reality. Because of these new climate trends, scarcity of water resources presents the most threatening risk and its effects are beginning to affect several regions of the planet. In this regard, it cannot be denied that Morocco, due to its location in the south of the Mediterranean, has been affected by these climatic trends, of which the plateau de Sais offers a good example by its location in the center of the country and by its wealth of resources. superficial and underground. In recent decades, this plateau has experienced water stress, particularly in its underground water reserves, it has been manifested by the decrease in flows from several sources or even total drying up, in particular after successive periods of unusual drought which marked the whole of the country after the 1980s. It is in this perspective this study aims to understand the behavior of the hydrogeological functioning of the sources located in the plain of Fez and its middle Atlas borders. Given its geological complexity, as well as the abundance of carbonate rocks and the presence of volcanic basalts flowing from south to north, have favored the resurgence of several sources of different flows and with a hydrological regime influenced according to the years and seasons by rainfall inputs. All these physical factors as well as human intervention on the hydrosystem through hydro-agricultural activities influence the behavior of the hydrological functioning of these sources.
The EPIK method was the first groundwater vulnerability mapping method specifically dedicated to karst aquifers. It has now been used for more than twenty years for the delineation of karst groundwater protection zones in Switzerland. This approach, based on the mapping of four parameters (i.e. four sub-systems), including epikarst (E), protective cover (P), infiltration conditions (I) and karst network (K), contributed to improve land management and reduce some water quality problems.
Nevertheless, based on the experience related to the delineation of protection zones in many karst basins and on recent research focused on flow, transport and vulnerability assessment in karst, the necessity of updating the method was highlighted.
While the EPIK conceptual model and the four mapping parameters remain the same, some concepts have been adapted or newly implemented within the frame of a revised method:
better definition of the vulnerability assessment conditions by precising water quality objectives, hydrological conditions, and recharge scenario
more detailed integration of the flow and contaminant transport specificities linked to the four sub-systems according to a comprehensive origin-pathway-target approach
more quantitative criterion for the mapping and point counting of the different parameters and their combination to obtain vulnerability maps
development of a normalized tracer methodology to calibrate and validate vulnerability assesment
More precisely the following upgrades have been applied to the method:
Concerning the E and P parameters, classification and rating have been modified to better reflect the storage capacity which critically determine the protective function of the unsaturated zone. Besides, a more detailed protective cover mapping based on three ranges of hydraulic conductivities has been implemented. The mapping of epikarst includes from now a more conservative point attribution and validation procedure in absence of specific karstic geomorphology.
The influence of infiltration conditions is taken into account differently by lowering or suppressing the protective effect of E + P, function of the amount of meteoric recharge (diffuse, concentrated). Surface flow, i.e. runoff, is only mapped within the contribution area feeding concentrated recharge points.
The protective effect of karstic network / saturated karst is determined based on groundwater residence time in medium to high water stage conditions. The protection linked to this parameter is generally very low but can become significant in case of large catchment area or even dominant in presence of less karstified aquifer and for deep karst systems. On the contrary to the former EPIK method, vulnerability can now decrease function of the distance to the spring or well in some instances.
The combination of the four parameters allows first determining a vulnerability map and second delineating the groundwater protection zone according to Swiss legislation. The implementation of new guidelines concerning activities in the protection zones that considers the specific requirements of karst environments further improves the adequateness of the approach proposed.
Springwater is often perceived as natural and clean. However, this is not always the case, and especially not in karst. Here, we study 2 small karst springs from Northern Norway used for drinking water. These springs are not sampled or monitored by local authorities but are nevertheless appreciated by the local population. The karst aquifers are developed in marble stripe karst, where the primary porosity of the crystalline rock is negligible, and water storage and transport are restricted to secondary and third porosity. The development of karst conduits depends on secondary porosity such as fractures, faults, shear planes and rock interfaces. The marble bands contain relict cave systems of several kilometres in length. The vadose stream sections are short and has an insignificant imprint on the cave morphology and as such has an invasion character in the pre-existing cave systems. Consequently, the positioning and size of the cave passages suggest speleogenesis (sensu lato) in different climatic or topographical setting, i.e. a subglacial hydrological regime. In contrast, the recent karst aquifers seem to be dominated by smaller conduits and submerge the lower parts of the existing cave systems to a minor extent.
At present, the recharge to the karst aquifers is dominated by diffuse infiltration. We generated vulnerability maps applying the EPIK method that considers the epikarst, protection cover, infiltration condition, and karst network development and combined them with geochemical monitoring of the springs. Our results show that most of the catchment areas are moderate to highly vulnerable because of the thin soil cover and the abundance of small surface karstic depressions. Water analyses demonstrate that the water is clean regarding pathogens and hazardous chemical constituents under normal flow conditions. However, climate change with more intense precipitation and prolonged dry periods may stress these hydrological systems. Monitoring precipitation and drip rates in the caves combined with physical and chemical characteristics in the springs will provide insights into the hydrodynamic and vulnerability of the aquifers.
The edge zone of sedimentary basins is the site of interactions involving the recharge of multi-layer aquifers, affecting the quality of the water resource. At the regional scale, the karst feature of the reservoirs is rarely taken into account in hydrogeological processes. However, karst makes complex the recharge, the exchanges and the mixing between reservoirs.
This study aims to characterize the regional hydrodynamic and hydrochemical behavior of multi-layer carbonate aquifers located in the basin edge by considering their karst feature.
Thereby, a dataset of water level and physico-chemical analyses in 66 wells and 73 springs was collected from French public data in the Northern Aquitaine basin, in France. A statistical approach was conducted in order to differentiate water compartments in relation with their hydrogeological context. First results show a contrasted physico-chemical response between Jurassic and Cretaceous aquifers, which is explained by the lithological properties, the aquifer depth, and the residence time. Nevertheless, some temperature and hydrochemical anomalies are identified in some sites. Likely hypothesis are the karst influence due to localized recharge from river losses, as well as the karstification degree in unconfined and confined compartments.
To validate theses hypothesis, current works are performed to compare i) the hydraulic response of 19 pumping wells, and ii) the geochemical signal of 43 key sites using a multi-tracer approach combining tracers of reservoir (87Sr/86Sr, Mg), of recharge (18O/16O and 2H/1H), of residence time (13C/14C), of flow conditions (pCO2_ISc), and of contamination (CO, 18O-NO3 vs 15N, 11B/10B). The expected results relate to the definition of a regional conceptual hydrogeological model of multi-layer aquifers along the edge of a sedimentary basin.
Tracer tests are widely used for characterizing hydrodynamics, from stream scale to basin-wide scale. In karstic environments, the positioning of field fluorometers (or sampling) is mostly determined by on-site configuration and setup difficulties. Most users are probably aware of the importance of this positioning for the relevance of data, and single-point tests are considered reliable. However, this importance is subjective to the user and the impact of positioning is not well quantified. Therefore, this study aims at quantifying the spatial heterogeneity of tracer concentration through time in karstic environment, and its impact on tracer test results and derived information on local hydrodynamics. The use of multi-point dye tracing is appropriate in the scope of this study.
Multi-point dye tracing consists in the use of multiple fluorometers for one tracer injection which can be placed in various configurations. Three major fluorometers configurations can be considered: transversal multi-point, along-stream multi-point and capacitive zone multi-point. (i) Transversal is the placement of multiple fluorometers across one river cross-section; (ii) along-stream considers a longitudinal placement of the fluorometers at various distances from the injection point; (iii) capacitive zone multi-point is the placement of fluorometers all around a large body of water such as a lake. Each of these configurations can help to characterize and quantify the influence of various hydrodynamical phenomena on the dispersion of solute.
Early results from transversal configuration in Wallonia (Belgium) indicate significant heterogeneity of tracer concentration across a stream section through time. These heterogeneities seem to be well correlated with anomalies in the conduit geometry, thus indicating that local karst geometry influences spatial distribution of tracer through time, and induces heterogeneity of concentration along a cross-section. The anomalies of a conduit geometry can induce various hydrodynamical features such as eddies, reverse current, slow or dead zone, turbulence, stream split... Further investigations on various sites with different karst geometry are considered in this project. Results from recent multi-point tracer test(s) will be presented.
Notable delay of more than one year (12.5–15.9 months) between precipitation recharge and groundwater-level upturn was recorded in the Triassic karstic aquifer of the Kopa Mt. (1187 m a.s.l.) in the middle of the West Carpathians in Slovakia. The velocity of recharge transit (0.24–0.70 m/day) was affirmed by borehole hydrographs. The groundwater level here is deep (113–300 m below ground surface) but approximately evenly distributed at the same elevation (~467 ± 10 m a.s.l.). Dolomites extend over 78% of the surface area of the karstic reservoir, with a Cretaceous marly aquitard boundary on the perimeter and on its base. The whole area is recharged merely by precipitation and entirely discharged by springs and small surface streams. The water balance between recharged and discharged water, and the groundwater-level fluctuation (shift in time), enabled calculation of stored groundwater volumes. The estimated specific yield of the Triassic dolomitic aquifer was unusually high (7.6–23.4%, mean 16.1%). Karstic cavities were also widely found in the borehole logs of nine boreholes that were 87–490-m deep and drilled in Triassic dolomites and limestones along the axis of a planned highway tunnel. Groundwater level monitoring took place over 40 months, between June 2014 and October 2017. Typical seasonal water-level oscillations were absent in the karstic reservoir, but water-level change was relevant to delayed past recharge events. In a borehole situated in the Cretaceous marly aquitard, good seasonal correlation with effective precipitation and also tidal response was observed.