Karst Catchment Research Articles

Unravelling nitrate transformation mechanisms in karst catchments through the coupling of high-frequency sensor data and machine learning

Nitrate dynamics within a catchment are critical to the earth's system process, yet the intricate details of its transport and transformation at high resolutions remain elusive. Hydrological effects on nitrate dynamics in particular have not been thoroughly assessed previously and this knowledge gap hampers our understanding and effective management of nitrogen cycling in watersheds. Here, machine learning (ML) models were employed to reconstruct the annual variation trend in nitrate dynamics and isotopes within a typical karst catchment. Random forest model demonstrates promising potential in predicting nitrate concentration and its isotopes, surpassing other ML models (including Long Short-term Memory, Convolutional Neural Network, and Support Vector Machine) in performance. The ML-modeled NO3−-N concentrations, δ15N-NO3−, and δ18O-NO3− values were in close agreement with field data (NSE values of 0.95, 0.80, and 0.53, respectively), which are notably challenging to achieve for process models. During the transition from dry to wet period, approximately 23.0% of the annual precipitation (∼269.1 mm) was identified as the threshold for triggering a rapid response in the wet period. The modeled nitrate isotope values were significantly supported by the field data, suggesting seasonal variations of nitrogen sources, with precipitation as the primary driving force for fertilizer sources. Mixing of multiple sources appeared to be the main control of the transport and transformation of nitrate during the rising limb in the wet period, whereas process control (denitrification) took precedence during the falling limb, and the fate of nitrate was controlled by biogeochemical processes during the dry period.

Effects of model complexity on karst catchment runoff modeling for flood warning systems

Effects of model complexity on karst catchment runoff modeling for flood warning systems

Machine Learning-based Optimal Nitrate Dynamics and Flux Monitoring Frequency in Karst Catchment

Machine Learning-based Optimal Nitrate Dynamics and Flux Monitoring Frequency in Karst Catchment

Can temperature be a low-cost tracer for modelling water age distributions in a karst catchment?

To investigate the feasibility of using temperature for tracking rainfall-runoff processes in karst catchments, this study developed a tracer-aided conceptual model using temperature as a tracer by coupling water and heat transport processes at the catchment scale. The model was calibrated and validated using hourly hydrometeorological and temperature data from a 1.25 km2 karst catchment in south-western China. The results showed that the model was able to capture the water flux and temperature dynamics of different landscape units in the karst catchment. Utilizing this framework, the model delineated the flux age distribution within different landscape units, as well as the overall water transit times through the catchment. The average flux ages were determined to be approximately 80 days for the hillslope unit, 452 days for the slow flow system, and 260 days for the fast flow regime within the depression areas. These estimations align broadly with those acquired using stable isotopes as tracers. Comparative analysis revealed that the flux age distributions derived from both temperature and isotopic tracers exhibited analogous patterns at the catchment outlet and across the hillslope compartments. However, the simulations based on temperature hinted at a heightened proportion of exceedingly young and decidedly old water in the outflow, alluding to a potential overestimation of these extreme age classes by the temperature-tracer model. From the temperature-simulated transit time distribution, about 31 % of the precipitation entering during the study period have left the catchment within 3 years, and a notable proportion of rain water was either stored in the aquifer or lost through evapotranspiration. The general characteristics of the transit time distribution simulated using temperature was similar with that simulated using isotopes, though a higher proportion of precipitation being drained by fast flows was inferred from the transit time distribution simulated using temperature. Collectively, our study demonstrated that temperature can serve as a cost-effective tracer for modelling of water age distributions and associated hydrological processes in karst catchments.

Predominant biogenic contribution of dissolved inorganic carbon in karst rivers, Southwest China

Dissolved inorganic carbon (DIC) holds significant importance within river systems, particularly in karst regions. The concentration of DIC is intricately linked to both biological processes, such as the respiration of organic matter, and geological processes, notably the strength of carbonate weathering. This dual influence exerts a profound impact on regional and global carbon cycles. However, the precise biological and geological origins of DIC in karst rivers remain a subject of debate. In this study, we delineate the sources of DIC in a typical karst catchment (Houzhai catchment) through isotopic analysis coupled with water chemistry examination. The results showed that δ13CDIC and Δ14CDIC variations range from −14.0 ‰ to −7.3 ‰ and − 426 ‰ to −31 ‰, respectively, with average values of −10.5 ‰ and − 180 ‰. Results derived from the isotopic mixing model suggest that approximately 83 % of DIC emanates from the soil CO2, while 17 % originates from carbonate sources. This indicates that biological processes exert a greater influence on DIC within the study area compared to geological processes. Seasonal variation, exemplified by higher DOC concentrations and Δ14CDIC values, alongside with lower DIC concentrations and δ13CDIC values during the wet season indicate increased input from biogenic sources, closely related to temperature and moisture conditions. Our study underscores an unbalanced contribution from biogenic and geogenic sources in karst rivers, highlighting potential inaccuracies in past carbon sink flux calculations. The roles of biogenic and geogenic sources in the carbon cycling of karst rivers, as well as their responses to future climate change, warrant a reevaluation.

Assessing current and future available resources to supply urban water demands using a high-resolution SWAT model coupled with recurrent neural networks and validated through the SIMPA model in karstic Mediterranean environments

Hydrological simulation in karstic areas is a hard task due to the intrinsic intricacy of these environments and the common lack of data related to their geometry. Hydrological dynamics of karstic sites in Mediterranean semiarid regions are difficult to be modelled mathematically owing to the existence of short wet episodes and long dry periods. In this paper, the suitability of an open-source SWAT method was checked to estimate the comportment of a karstic catchment in a Mediterranean semiarid domain (southeast of Spain), which wet and dry periods were evaluated using box-whisker plots and self-developed wavelet test. A novel expression of the Nash–Sutcliffe index for arid areas (ANSE) was considered through the calibration and validation of SWAT. Both steps were completed with 20- and 10-year discharge records of stream (1996–2015 to calibrate the model as this period depicts minimum gaps and 1985–1995 to validate it). Further, SWAT assessments were made with records of groundwater discharge and relating SWAT outputs with the SIMPA method, the Spain’s national hydrological tool. These methods, along with recurrent neural network algorithms, were utilised to examine current and predicted water resources available to supply urban demands considering also groundwater abstractions from aquifers and the related exploitation index. According to the results, SWAT achieved a “very good” statistical performance (with ANSE of 0.96 and 0.78 in calibration and validation). Spatial distributions of the main hydrological processes, as surface runoff, evapotranspiration and aquifer recharge, were studied with SWAT and SIMPA obtaining similar results over the period with registers (1980–2016). During this period, the decreasing trend of rainfalls, characterised by short wet periods and long dry periods, has generated a progressive reduction of groundwater recharge. According to algorithms prediction (until 2050), this declining trend will continue reducing groundwater available to meet urban demands and increasing the exploitation index of aquifers. These results offer valuable information to authorities for assessing water accessibility and to provide water demands in karstic areas.

High sensitivity of dissolved organic carbon transport during hydrological events in a small subtropical karst catchment

Dissolved organic carbon (DOC) and discharge are often tightly coupled, though these relationships in karst environments remain poorly constrained. In this study, DOC dynamics over 13 hydrological events, alongside monthly monitoring over an entire hydrological year were monitored in a small karst catchment, SW China. The concurrent analyses of power-law model and hysteresis patterns reveal that DOC behavior is generally transport-limited due to flushing effects of increased discharge but highly variable at both intra- and inter-event scales. The initial discharge at event onset and discharge-weighted mean concentration of DOC ([DOC]DW) of individual events can explain 37.7 % and 19.9 % of the variance of DOC behavior among events, respectively. The sustained dry-cold antecedent conditions make DOC hysteresis behavior during the earliest event complex and different from subsequent events. At event scale, the variability in DOC export is primarily controlled by [DOC]DW (explaining 64.3 %) and the yield of total dissolved solutes (YTDS, explaining 30.4 %), reflecting the impacts of variable hydrological connectivity and intense soil-water-rock interactions in this karst catchment. On an annual scale, DOC yield (YDOC, 222.86 kg C km−2) was mostly derived during the wet season (98.19 %) under the hydrological driving force. The difference in annual YDOC between this karst catchment and other regions can be well explained by annual water yield (Ywater, explaining 24.2 %) and [DOC] (explaining 35.4 %), whereas the variance in DOC export efficiency among catchments is almost exclusively controlled by [DOC] alone, independent of drainage area and annual Ywater. This study highlights the necessity of high-frequency sampling for modeling carbon biogeochemical processes and the particularity of the earliest hydrological events occurred after a long cold-dry period in karst catchments. Under the changing climate, whether DOC dynamics in karst catchments will present source-limited patterns during more extreme hydrological events merits further study.

Hydrochemical characterization of a Dinaric karst catchment in relation to emerging organic contaminants

The main findings of a hydrochemical investigation conducted within a typical Dinaric karst catchment located in Southern Croatia are outlined. The studied aquifer is drained by the Jadro and Žrnovnica springs, which are important for the regional and local water supplies, respectively. Presumably, there is intercatchment groundwater flow coming from the neighbouring Cetina River catchment. Various factors governing aquifer hydrochemistry and their interplay with emerging organic contaminants (EOCs) that were detected at different water resources in ng/L concentrations was assessed. A total of 26 sampling campaigns (October 2019 – October 2022) were conducted at two springs, in a river and at a deep borehole, all representative of this complex hydrogeological system. Assessment of major ion constituents and saturation indices calculated with PHREEQC revealed the sampled water resources are of a Ca-HCO3 type due to the predominant weathering of the carbonate mineral calcite. Sharp spikes observed in chemographs indicated a highly karstified system with an effective conduit network allowing rapid spring responses to precipitation events. Water resources are of good chemical status, as affirmed by anthropogenic contamination indicators, with nitrates, chlorides and sulphates all below maximum threshold values. Strong positive correlations were found between EOCs concentration, number of detected compounds, and nitrates in the Cetina River, indicating a common origin, most likely wastewater. Identification of persistent EOCs including widely used repellent N,N-diethyl- metatoluamide (DEET) during base flow conditions and its strong positive correlation with the Ca2+ content in both the Cetina and Jadro samples, suggests potential storage in the epikarst and aquifer matrix. This coupling of conventional hydrochemical indicators and novel markers of anthropogenic impacts, including EOCs, in vulnerable karst water resources is a crucial advancement in the assessment and management of emerging environmental and potential human health risks. Such an approach is pivotal for the sustainable protection of hydrogeologically intricate sites.

From Recharge to Cave to Spring: Transmission of a Flood Pulse through a Complex Karst Conduit Network, Castleton, Derbyshire (UK)

Storm Babet (18–21 October 2023) brought heavy and persistent rain (80–100 mm) to the English Peak District, causing widespread surface and underground flooding. The village of Castleton experienced groundwater flooding from springs that drain a complex mixed allogenic–autogenic karst catchment. Transmission of the flood pulse was monitored using high-resolution (2 and 4 min intervals) logging of (a) the hydraulic head at five underground locations in the karst conduits and (b) the water depth at three springs and in the surface river fed by the springs. Underground, there were large increases in the hydraulic head (9–35 m), which resulted in two types of flow switching. Firstly, the increased head at the input end of a phreatic (water-filled) conduit system removed an underwater permeability barrier in a relatively low-elevation conduit, resulting in a dramatic increase in flow out of the conduit and a corresponding decrease in flow from a linked higher-elevation conduit that had dominated before the storm. Secondly, the increased head upstream of two conduits with limited hydraulic conductivity allowed water to spill over into conduits that were inactive prior to the storm. As expected, the conduits fed by sinking streams from the allogenic catchment responded rapidly to the recharge, but there was also a rapid response from the autogenic catchment where there are no surface streams and only a small number of dolines. The complex signals measured underground are not apparent from the spring hydrographs.

Response of weathering carbon sink effect to anthropogenic sulfuric acid in different lithological catchments: A case study from Southwest China

Accurately quantifying the carbon sink effect resulting from chemical weathering caused by anthropogenic H2SO4 is imperative to improve the assessment of the global carbon budget. Nevertheless, there is still a lack of precise understanding regarding the impact of anthropogenic H2SO4 on CO2 consumption during chemical weathering. Here, spring water samples were collected monthly from three catchments with distinct bedrock lithologies affected by severe acid precipitation in Southwest China for analyses of hydrogeochemistry and δ13CDIC to quantitatively estimate the effect of anthropogenic H2SO4 on the weathering carbon sink budget. The results show that carbonates contribute 97.4 %, 95.0 % and 88.8 % of the total cationic load using a straightforward method in the Beidiping carbonate catchment, as well as in the Shegengyan and Bianyan silicate catchments, respectively. The [Ca2++Mg2+]/[HCO3−] (0.98–1.19) and [Ca2++Mg2+]/[HCO3−+SO42−] (approximately 1) equivalent ratios, and δ13CDIC values (−16.8 to −8.0 ‰) of the samples suggest that besides H2CO3, H2SO4 is involved in carbonate weathering. The stoichiometry of the chemical compositions of spring water indicates that the presence of H2SO4 enhances carbonate weathering rates by 14.8 %, 8.1 % and 7.5 % while decreasing the CO2 consumption by 8.2 %, 4.3 % and 4.0 % in Beidiping, Shegengyan and Bianyan, respectively. Thus the reduced proportion of karst carbon sink in the carbonate catchment is approximately 2 times that in the silicate catchment, suggesting that carbonate weathering in the karst catchment is more sensitive to acid precipitation. The impact of acid precipitation on rock weathering in the silicate catchment is constrained by the soil buffering effect. Our study highlights the important role of anthropogenic H2SO4 in carbonate weathering, which should be critically evaluated in regional and global carbon cycles in future studies.

Combustion Completeness and Sample Location Determine Wildfire Ash Leachate Chemistry

AbstractUnderstanding past fire regimes and how they vary with climate, human activity, and vegetation patterns is fundamental to the mitigation and management of changing fire regimes as anthropogenic climate change progresses. Ash‐derived trace elements and pyrogenic biomarkers from speleothems have recently been shown to record past fire activity in speleothems from both Australia and North America. This calls for an empirical study of ash geochemistry to aid the interpretation of speleothem palaeofire proxy records. Here we present analyses of leached ashes collected following fires in southwest and southeast Australia. We include a suite of inorganic elemental data from the water‐soluble fraction of ash as well as a selection of organic analytes (pyrogenic lipid biomarkers). We also present elemental data from leachates of soils collected from sites in southwest Australia. We demonstrate that the water‐soluble fraction of ash differs from the water‐soluble fraction of soils, with trace and minor element concentrations in ash leachates varying with combustion completeness (burn severity) and sample location. Changes in some lipid biomarker concentrations extracted from ashes may reflect burn severity. Our results contribute to building a process‐based understanding of how speleothem geochemistry may record fire frequency and severity, and suggest that more research is needed to understand the transport pathways for the inclusion of pyrogenic biomarkers in speleothems. Our results also demonstrate that potential contaminant loads from ashes are much higher than from soils, with implications for the management of karst catchments, which are a critical water resource.

Differences in soil water movement between the dip and anti-dip slopes of a karst trough valley

Exploring temporal and spatial variations in soil water processes is crucial for identifying the ecological water supply capacity of soil and understanding hydrologic cycles within karst catchments. However, the existing body of knowledge related to soil water dynamics across dip and anti-dip slopes in karst landscapes remains limited. To address this gap, this study investigated soil water dynamics in a karst trough valley (KTV) in southwest China, using stable isotopes (δD and δ18O) from eight sampling sites and precipitation data from April 2017 to November 2018. Our findings suggest that seasonal variations in soil water isotopes are primarily influenced by climatic factors, with soil water lagging behind rainwater isotopes by one to four months due to differences in the residence time of precipitation in different soil layers. Spatially, our results reveal the coexistence of diffuse and preferential flows on both dip and anti-dip slopes. Specifically, on the dip slope, preferential flow dominates above 20 cm, while diffuse flow prevails below this depth. In contrast, on the anti-dip slope, preferential flow is prominent above 40 cm depth, while diffuse flow dominates beneath this threshold. These spatial patterns not only impact the ecological water and nutrient supply capacity of soil but they also contribute to distinct patterns in vegetative growth on both slopes. In turn, vegetation growth facilitates preferential flow pathways, establishing a positive feedback loop between soil water movement and vegetative cover. These findings may lead to a deeper understanding of soil water processes on the dip and anti-dip slopes of KTVs and provide crucial insights into the relationship between soil water and vegetative growth in similar karst landforms.

Groundwater hydrochemical signatures, nitrate sources, and potential health risks in a typical karst catchment of North China using hydrochemistry and multiple stable isotopes.

Nitrate pollution in aquatic ecosystems has received growing concern, particularly in fragile karst basins. In this study, hydrochemical compositions, multiple stable isotopes (δ2H-H2O, δ18Ο-Η2Ο, δ15Ν-ΝΟ3-, and δ18Ο-ΝΟ3-), and Bayesian stable isotope mixing model (MixSIAR) were applied to elucidate nitrate pollution sources in groundwater of the Yangzhuang Basin. The Durov diagram identified the dominant groundwater chemical face as Ca-HCO3 type. The NO3- concentration ranged from 10.89 to 90.45mg/L (average 47.34mg/L), showing an increasing trend from the upstream forest and grassland to the downstream agricultural dominant area. It is worth noting that 47.2% of groundwater samples exceeded the NO3- threshold value of 50mg/L for drinking water recommended by the World Health Organization. The relationship between NO3-/Cl- and Cl- ratios suggested that most groundwater samples were located in nitrate mixed endmember from agricultural input, soil organic nitrogen, and manure & sewage. The Self-Organizing Map (SOM) and Pearson correlations analysis further indicated that the application of calcium fertilizer, sodium fertilizer, and livestock and poultry excrement in farmland elevated NO3- level in groundwater. The output results of the MixSIAR model showed that the primary sources of NO3- in groundwater were soil organic nitrogen (55.3%), followed by chemical fertilizers (28.5%), sewage & manure (12.7%), and atmospheric deposition (3.4%). Microbial nitrification was a dominant nitrogen conversion pathway elevating NO3- levels in groundwater, while the denitrification can be neglectable across the study area. The human health risk assessment (HHRA) model identified that about 88.9%, 77.8%, 72.2%, and 50.0% of groundwater samples posing nitrate's non-carcinogenic health hazards (HQ > 1) through oral intake for infants, children, females, and males, respectively. The findings of this study can offer useful biogeochemical information on nitrogen pollution in karst groundwater to support sustainable groundwater management in similar human-affected karst regions.

Improving the simulations of the hydrological model in the karst catchment by integrating the conceptual model with machine learning models

Hydrological modelling can be complex in nonhomogeneous catchments with diverse geological, climatic, and topographic conditions. In this study, an integrated conceptual model including the snow module with machine learning modelling approaches was implemented for daily rainfall-runoff modelling in mostly karst Ljubljanica catchment, Slovenia, which has heterogeneous characteristics and is potentially exposed to extreme events that make the modelling process more challenging and crucial. In this regard, the conceptual model CemaNeige Génie Rural à 6 paramètres Journalier (CemaNeige GR6J) was combined with machine learning models, namely wavelet-based support vector regression (WSVR) and wavelet-based multivariate adaptive regression spline (WMARS) to enhance modelling performance. In this study, the performance of the models was comprehensively investigated, considering their ability to forecast daily extreme runoff. Although CemaNeige GR6J yielded a very good performance, it overestimated low flows. The WSVR and WMARS models yielded poorer performance than the conceptual and hybrid models. The hybrid model approach improved the performance of the machine learning models and the conceptual model by revealing the linkage between variables and runoff in the conceptual model, which provided more accurate results for extreme flows. Accordingly, the hybrid models improved the forecasting performance of the maximum flows up to 40 % and 61 %, and minimum flows up to 73 % and 72 % compared to the CemaNeige GR6J and stand-alone machine learning models. In this regard, the hybrid model approach can enhance the daily rainfall-runoff modelling performance in nonhomogeneous and karst catchments where the hydrological process can be more complicated.

Characterizing occurrence of emerging organic contaminants in Dinaric karst catchment of Jadro and Žrnovnica springs, Croatia

The occurrence of emerging organic contaminants (EOCs) was investigated in vulnerable Dinaric karst catchment of Jadro and Žrnovnica springs in Croatia, under varying flow conditions and across three different water resource types (groundwater, springs, and surface water). The maximal EOCs concentration in both springs were observed following autumn recharge events, while during peak discharge no detection above the limits of detection (LOD) was recorded due to dilution process. Contrarily to springs, groundwater from deep borehole exhibited highest total EOCs concentration under low flow conditions, underscoring the considerable karst aquifer vulnerability and its oligotrophic nature. The peak EOCs concentration in karstic Cetina River coincided with the river's lowest discharge. The highest mass flux of 1013 g/day was determined for very mobile pharmaceutical metformin detected in Cetina. The presence of potentially persistent to very persistent compounds, like DEET and 1H-benzotriazole, which exhibited highest detection frequencies across all sampling sites, was observed in association with varying hydrological conditions. Hypotheses regarding the occurrence of identified EOCs include surface contamination infiltrating directly through ponors and highly karstified areas, potential persistence in the epikarst and aquifer matrix, and site-specific contamination sources for compounds such as 1H-benzotriazole, gabapentin, and ketoprofen found in groundwater. To evaluate the risk of inadvertent human exposure to EOCs across various age groups, we utilized measured spring concentrations and calculated drinking water equivalent levels (DWEL), which ranged from 1.4 μg/L for cotinine to 503 mg/L for sucralose, both detected in Jadro spring. Although EOCs concentrations in ng/L are unlikely to pose a significant risk to healthy population, long-term exposure to EOCs mixture remains unknown. Given scarcity of research on EOCs in karst environments on both global and national levels, our study enhances comprehension of their occurrence and behaviour across different karst water resources that hold crucial importance for drinking water supply in regions like Dinarides.

Human-environment interactions in two Karst catchments – Gradole (NW Croatia) and Ljubija springs (NE Slovenia)

Human-environment interactions in two Karst catchments – Gradole (NW Croatia) and Ljubija springs (NE Slovenia)

Amplified Production and Export of Dissolved Inorganic Carbon During Hot and Wet Subtropical Monsoon

AbstractUnderstanding the origins and processes of riverine dissolved inorganic carbon (DIC) is crucial for predicting the global carbon cycle with projected, more frequent climate extremes yet our knowledge has remained fragmented. Here we ask: How and how much do DIC production and export vary across space (shallow vs. deep, uphill vs. depression) and time (daily, seasonal, and annual)? How do the relative contributions of biogenic (soil respiration) and geogenic (carbonate weathering) sources differ under different temperature and hydrological conditions? We answer these questions using a catchment‐scale reactive transport model constrained by stream flow, stable water isotopes, stream DIC, and carbon isotope data from a headwater karstic catchment in southwest China in a subtropical monsoon climate. Results show climate seasonality regulates the timing of DIC production and export. In hot‐wet seasons, high temperature accelerates soil respiration and carbonate weathering (up to a factor of three) via elevating soil CO2 and carbonate solubility, whereas high discharge enhances export by two orders of magnitude compared to cold‐dry seasons. Carbonate weathering is driven more by soil CO2 than water flow. At the annual scale, 92.9% and 7.1% of DIC was produced in shallow and deep zone, respectively, whereas 64.5% and 35.5% of DIC was exported from shallow and deep zone, respectively. These results highlight the uniqueness of subtropical karst areas as synchronous reactors and transporters of DIC during the hot‐wet monsoon, contrasting the asynchronous production and export in other climate regions. A future hotter and wetter climate with more intensive storms in the region may further intensify DIC production and export, accentuating the potential of subtropical karst regions as global hot spots for carbon cycling.

Improving understanding of groundwater flow in an alpine karst system by reconstructing its geologic history using conduit network model ensembles

Abstract. Reconstructing the geologic history of a karst area can advance understanding of the system's present-day hydrogeologic functioning and help predict the location of unexplored conduits. This study tests competing hypotheses describing past conditions controlling cave formation in an alpine karst catchment, by comparing an ensemble of modeled networks to the observed network map. The catchment, the Gottesacker karst system (Germany and Austria), is drained by three major springs and a paleo-spring and includes the partially explored Hölloch cave, which consists of an active section whose formation is well-understood and an inactive section whose formation is the subject of debate. Two hypotheses for the formation of the inactive section are the following: (1) glaciation obscured the three present-day springs, leaving only the paleo-spring, or (2) the lowest of the three major springs (Sägebach) is comparatively young, so its subcatchment previously drained to the paleo-spring. These hypotheses were tested using the pyKasso Python library (built on anisotropic fast-marching methods) to generate two ensembles of networks, one representing each scenario. Each ensemble was then compared to the known cave map. The simulated networks generated under hypothesis 2 match the observed cave map more closely than those generated under hypothesis 1. This supports the conclusion that the Sägebach spring is young, and it suggests that the cave likely continues southwards. Finally, this study extends the applicability of model ensemble methods from situations where the geologic setting is known but the network is unknown to situations where the network is known but the geologic evolution is not.

Hourly rainfall-runoff modelling by combining the conceptual model with machine learning models in mostly karst Ljubljanica River catchment in Slovenia

Hydrological modelling, essential for water resources management, can be very complex in karst catchments with different climatic and geologic characteristics. In this study, three combined conceptual models incorporating the snow module with machine learning models were used for hourly rainfall-runoff modelling in the mostly karst Ljubljanica River catchment, Slovenia. Wavelet-based Extreme Learning Machine (WELM) and Wavelet-based Regression Tree (WRT) machine learning models were integrated into the conceptual CemaNeige Génie Rural à 4 paramètres Horaires (CemaNeige GR4H). In this regard, the performance of the hybrid models was compared with stand-alone conceptual and machine learning models. The stand-alone WELM and WRT models using only meteorological variables performed poorly for hourly runoff forecasting. The CemaNeige GR4H model as stand-alone model yielded good performance; however, it overestimated low flows. The hybrid CemaNeige GR4H-WELM and CemaNeige-WRT models provided better simulation results than the stand-alone models, especially regarding the extreme flows. The results of the study demonstrated that using different variables from the conceptual model, including the snow module, in the machine learning models as input data can significantly affect the performance of rainfall-runoff modelling. The hybrid modelling approach can potentially improve runoff simulation performance in karst catchments with diversified geological formations where the rainfall-runoff process is more complex.

Impact of land-use change on karst spring response by integration of surface processes in karst hydrology: The ISPEEKH model

Karst groundwater represents a source of freshwater for a quarter of the world's population. Modeling the hydrological behavior of karst aquifers while accounting for the spatial variability of the recharge processes and the land-use change impacts on water availability remains a challenge for karst water resources management. Therefore, this study proposes a new semi-distributed hydrological model named Integration of Surface ProcEssEs in Karst Hydrology (ISPEEKH) for simulating the surface-underground water flows in karst-dominated catchments. ISPEEKH couples the restructured version of the Soil and Water Assessment Tool (SWAT+) with a non-linear three-reservoir groundwater model that represents the epikarst and matrix-conduit flow components in a karst aquifer. ISPEEKH was applied to the Baget karst catchment (13.25 km2) in the Pyrénées mountains of southwest France, characterized by a conduit-dominated spring flow and low surface runoff. The model simulated the catchment daily discharge satisfactorily, with NSE = 0.67 and R2 = 0.68 for the 2008–2013 calibration period, and NSE = 0.65 and R2 = 0.69 for the 2014–2018 validation period. The catchment evapotranspiration and discharge amounted to 33 % and 67 % of the average annual precipitation, respectively, and nearly 80 % of the discharge was attributed to spring flow. The matrix-to-conduit flow accounted for 4–7 % of the monthly spring flow during the high-flow period (December to June), and up to 22 % during the low-flow period (July to October). ISPEEKH was then used to assess the hydrological response of the Baget catchment to land-use change scenarios of afforestation and deforestation for wood production and pasture development. Afforestation of the full catchment area did not alter the catchment hydrology significantly. Deforestation for wood production reduced annual evapotranspiration by 12–18 % and increased annual recharge by 7–8 %, resulting in annual discharge rising by 6–9 %. These changes in the water balance components were most noticeable during the low-flow season and would augment freshwater availability. Conversely, deforestation for pasture development raised annual evapotranspiration by 13–17 % while reducing annual recharge by around 7 % and discharge by 5–7 %. These changes in the hydrological components were most remarkable during the high-flow season and affected discharge below 2.5 m3.s−1.