Articles | Volume 29, issue 15
https://doi.org/10.5194/hess-29-3405-2025
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/hess-29-3405-2025
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
01 Aug 2025
Research article |
| 01 Aug 2025
| 01 Aug 2025
Towards a global spatial machine learning model for seasonal groundwater level predictions in Germany
Federal Institute for Geosciences and Natural Resources, Berlin, Germany
Alexander Schulz
Berliner Hochschule für Technik, Berlin, Germany
Maria Wetzel
Federal Institute for Geosciences and Natural Resources, Berlin, Germany
Maximilian Nölscher
Federal Institute for Geosciences and Natural Resources, Berlin, Germany
Teodor Chiaburu
Berliner Hochschule für Technik, Berlin, Germany
Felix Biessmann
Berliner Hochschule für Technik, Berlin, Germany
Stefan Broda
Federal Institute for Geosciences and Natural Resources, Berlin, Germany
Related authors
No articles found.
Marc Ohmer, Tanja Liesch, Bastian Habbel, Benedikt Heudorfer, Mariana Gomez, Patrick Clos, Maximilian Nölscher, and Stefan Broda
Earth Syst. Sci. Data Discuss., https://doi.org/10.5194/essd-2025-321, https://doi.org/10.5194/essd-2025-321, 2025
Preprint under review for ESSD
Short summary
Short summary
We present a public dataset of weekly groundwater levels from more than 3,000 wells across Germany, spanning 32 years. It combines weather data and site-specific environmental information to support forecasting groundwater changes. Three benchmark models of varying complexity show how data and modeling approaches influence predictions. This resource promotes open, reproducible research and helps guide future water management decisions.
Raoul A. Collenteur, Ezra Haaf, Mark Bakker, Tanja Liesch, Andreas Wunsch, Jenny Soonthornrangsan, Jeremy White, Nick Martin, Rui Hugman, Ed de Sousa, Didier Vanden Berghe, Xinyang Fan, Tim J. Peterson, Jānis Bikše, Antoine Di Ciacca, Xinyue Wang, Yang Zheng, Maximilian Nölscher, Julian Koch, Raphael Schneider, Nikolas Benavides Höglund, Sivarama Krishna Reddy Chidepudi, Abel Henriot, Nicolas Massei, Abderrahim Jardani, Max Gustav Rudolph, Amir Rouhani, J. Jaime Gómez-Hernández, Seifeddine Jomaa, Anna Pölz, Tim Franken, Morteza Behbooei, Jimmy Lin, and Rojin Meysami
Hydrol. Earth Syst. Sci., 28, 5193–5208, https://doi.org/10.5194/hess-28-5193-2024, https://doi.org/10.5194/hess-28-5193-2024, 2024
Short summary
Short summary
We show the results of the 2022 Groundwater Time Series Modelling Challenge; 15 teams applied data-driven models to simulate hydraulic heads, and three model groups were identified: lumped, machine learning, and deep learning. For all wells, reasonable performance was obtained by at least one team from each group. There was not one team that performed best for all wells. In conclusion, the challenge was a successful initiative to compare different models and learn from each other.
Mariana Gomez, Maximilian Nölscher, Andreas Hartmann, and Stefan Broda
Hydrol. Earth Syst. Sci., 28, 4407–4425, https://doi.org/10.5194/hess-28-4407-2024, https://doi.org/10.5194/hess-28-4407-2024, 2024
Short summary
Short summary
To understand the impact of external factors on groundwater level modelling using a 1-D convolutional neural network (CNN) model, we train, validate, and tune individual CNN models for 505 wells distributed across Lower Saxony, Germany. We then evaluate the performance of these models against available geospatial and time series features. This study provides new insights into the relationship between these factors and the accuracy of groundwater modelling.
Benedikt Heudorfer, Tanja Liesch, and Stefan Broda
Hydrol. Earth Syst. Sci., 28, 525–543, https://doi.org/10.5194/hess-28-525-2024, https://doi.org/10.5194/hess-28-525-2024, 2024
Short summary
Short summary
We build a neural network to predict groundwater levels from monitoring wells. We predict all wells at the same time, by learning the differences between wells with static features, making it an entity-aware global model. This works, but we also test different static features and find that the model does not use them to learn exactly how the wells are different, but only to uniquely identify them. As this model class is not actually entity aware, we suggest further steps to make it so.
Maria Wetzel, Thomas Kempka, and Michael Kühn
Adv. Geosci., 58, 1–10, https://doi.org/10.5194/adgeo-58-1-2022, https://doi.org/10.5194/adgeo-58-1-2022, 2022
Short summary
Short summary
Porosity-permeability relations are simulated for a precipitation-dissolution cycle in a virtual sandstone. A hysteresis in permeability is observed depending on the geochemical process and dominating reaction regime, whereby permeability varies by more than two orders of magnitude. Controlling parameters for this hysteresis phenomenon are the closure and re-opening of micro-scale flow channels, derived from changes in pore throat diameter and connectivity of the pore network.
Morgan Tranter, Maria Wetzel, Marco De Lucia, and Michael Kühn
Adv. Geosci., 56, 57–65, https://doi.org/10.5194/adgeo-56-57-2021, https://doi.org/10.5194/adgeo-56-57-2021, 2021
Short summary
Short summary
Barite formation is an important factor for many use cases of the geological subsurface because it may change the rock.
In this modelling study, the replacement reaction of celestite to barite is investigated.
The steps that were identified to play a role are celestite dissolution followed by two-step precipitation of barite: spontaneous formation of small crystals and their subsequent growth.
Explicitly including the processes improve the usability of the models for quantified prediction.
Andreas Wunsch, Tanja Liesch, and Stefan Broda
Hydrol. Earth Syst. Sci., 25, 1671–1687, https://doi.org/10.5194/hess-25-1671-2021, https://doi.org/10.5194/hess-25-1671-2021, 2021
Maria Wetzel, Thomas Kempka, and Michael Kühn
Adv. Geosci., 54, 33–39, https://doi.org/10.5194/adgeo-54-33-2020, https://doi.org/10.5194/adgeo-54-33-2020, 2020
Elena Tillner, Maria Langer, Thomas Kempka, and Michael Kühn
Hydrol. Earth Syst. Sci., 20, 1049–1067, https://doi.org/10.5194/hess-20-1049-2016, https://doi.org/10.5194/hess-20-1049-2016, 2016
Short summary
Short summary
The degree of shallow aquifer salinisation triggered by fluid injection into deeper brine-bearing aquifers and brine upward migration through hydraulically conductive faults strongly depends on the regional depth of the freshwater-saltwater boundary, since displaced brines originate only from the upper fault damage zones in the study area. The highest local salinity increase in shallow aquifers occurs in case of closed model boundaries and low fault damage zone volumes.
Related subject area
Subject: Groundwater hydrology | Techniques and Approaches: Modelling approaches
The impact of geological structures on groundwater potential assessment in volcanic rocks in the Borena Sayint district, northwestern Ethiopian Plateau: a review
Numerical analysis of the effect of heterogeneity on CO2 dissolution enhanced by gravity-driven convection
Multivariate and long-term time series analysis to assess the effect of nitrogen management policy on groundwater quality in Wallonia, BE
Laboratory heat transport experiments reveal grain-size- and flow-velocity-dependent local thermal non-equilibrium effects
Improvement of the KarstMod modelling platform for a better assessment of karst groundwater resources
Topothermohaline convection – from synthetic simulations to reveal processes in a thick geothermal system
Training deep learning models with a multi-station approach and static aquifer attributes for groundwater level simulation: what is the best way to leverage regionalised information?
Managed aquifer recharge and exploitation impacts on dynamics of groundwater level and quality in northern China karst area: Quantitative research by multi-methods
Integrated approach for characterizing aquifer heterogeneity in alluvial plains
Explicit simulation of reactive microbial transport with a dual-permeability, two-site kinetic deposition formulation using the integrated surface-subsurface hydrological model HydroGeoSphere
Improving heat transfer predictions in heterogeneous riparian zones using transfer learning techniques
Catchment landforms predict groundwater-dependent wetland sensitivity to recharge changes
Data-driven modelling of hydraulic-head time series: results and lessons learned from the 2022 Groundwater Time Series Modelling Challenge
The impact of future changes in climate variables and groundwater abstraction on basin-scale groundwater availability
Assessing groundwater level modelling using a 1-D convolutional neural network (CNN): linking model performances to geospatial and time series features
Reinforce lake water balance components estimations by integrating water isotope compositions with a hydrological model
Short high-accuracy tritium data time series for assessing groundwater mean transit times in the vadose and saturated zones of the Luxembourg Sandstone aquifer
High-resolution long-term average groundwater recharge in Africa estimated using random forest regression and residual interpolation
Towards understanding the influence of seasons on low-groundwater periods based on explainable machine learning
Shannon entropy of transport self-organization due to dissolution–precipitation reaction at varying Peclet numbers in initially homogeneous porous media
A high-resolution map of diffuse groundwater recharge rates for Australia
Influence of bank slope on sinuosity-driven hyporheic exchange flow and residence time distribution during a dynamic flood event
Technical note: A model of chemical transport in a wellbore–aquifer system
Disentangling coastal groundwater level dynamics in a global dataset
Current and future roles of meltwater–groundwater dynamics in a proglacial Alpine outwash plain
On the challenges of global entity-aware deep learning models for groundwater level prediction
Incorporating interpretation uncertainties from deterministic 3D hydrostratigraphic models in groundwater models
Adjoint subordination to calculate backward travel time probability of pollutants in water with various velocity resolutions
On the optimal level of complexity for the representation of groundwater-dependent wetland systems in land surface models
Estimation of groundwater age distributions from hydrochemistry: comparison of two metamodelling algorithms in the Heretaunga Plains aquifer system, New Zealand
Technical note: Novel analytical solution for groundwater response to atmospheric tides
Calibration of groundwater seepage against the spatial distribution of the stream network to assess catchment-scale hydraulic properties
Climate-warming-driven changes in the cryosphere and their impact on groundwater–surface-water interactions in the Heihe River basin
Comparison of artificial neural networks and reservoir models for simulating karst spring discharge on five test sites in the Alpine and Mediterranean regions
A general model of radial dispersion with wellbore mixing and skin effects
Estimation of hydraulic conductivity functions in karst regions by particle swarm optimization with application to Lake Vrana, Croatia
The origin of hydrological responses following earthquakes in a confined aquifer: insight from water level, flow rate, and temperature observations
Advance prediction of coastal groundwater levels with temporal convolutional and long short-term memory networks
Three-dimensional hydrogeological parametrization using sparse piezometric data
Machine-learning-based downscaling of modelled climate change impacts on groundwater table depth
Frequency domain water table fluctuations reveal impacts of intense rainfall and vadose zone thickness on groundwater recharge
Characterizing groundwater heat transport in a complex lowland aquifer using paleo-temperature reconstruction, satellite data, temperature–depth profiles, and numerical models
Karst spring recession and classification: efficient, automated methods for both fast- and slow-flow components
Exploring river–aquifer interactions and hydrological system response using baseflow separation, impulse response modeling, and time series analysis in three temperate lowland catchments
Experimental study of non-Darcy flow characteristics in permeable stones
Karst spring discharge modeling based on deep learning using spatially distributed input data
HESS Opinions: Chemical transport modeling in subsurface hydrological systems – space, time, and the “holy grail” of “upscaling”
Spatiotemporal variations in water sources and mixing spots in a riparian zone
Delineation of discrete conduit networks in karst aquifers via combined analysis of tracer tests and geophysical data
Reactive transport modeling for supporting climate resilience at groundwater contamination sites
Bishaw Mihret and Ajebush Wuletaw
Hydrol. Earth Syst. Sci., 29, 2951–2959, https://doi.org/10.5194/hess-29-2951-2025, https://doi.org/10.5194/hess-29-2951-2025, 2025
Short summary
Short summary
This review highlights how geological structures like faults and fractures influence groundwater potential in volcanic rocks of Borena Sayint, northwestern Ethiopian Plateau. These structures enhance secondary porosity, affecting aquifer behaviour. The study stresses integrating geological, geophysical, and hydrological data for sustainable groundwater management in tectonically complex volcanic regions.
Yufei Wang, Daniel Fernàndez-Garcia, and Maarten W. Saaltink
Hydrol. Earth Syst. Sci., 29, 2485–2503, https://doi.org/10.5194/hess-29-2485-2025, https://doi.org/10.5194/hess-29-2485-2025, 2025
Short summary
Short summary
During geological carbon sequestration, the injected supercritical CO2, being less dense, floats above the brine. The dissolution of CO2 into brine helps mitigate the risk of CO2 leakage. As CO2 dissolves into the brine, it increases the density of brine in the upper layer, initiating gravity-driven convection (GDC), which significantly enhances the rate of CO2 dissolution. We derived two empirical formulas to predict the asymptotic dissolution rate driven by GDC in heterogeneous fields.
Elise Verstraeten, Alice Alonso, Louise Collier, and Marnik Vanclooster
Hydrol. Earth Syst. Sci., 29, 1829–1845, https://doi.org/10.5194/hess-29-1829-2025, https://doi.org/10.5194/hess-29-1829-2025, 2025
Short summary
Short summary
This study takes a data-driven approach to evaluate long-term groundwater nitrate concentration trends in Wallonia, Belgium, over ~20 years following the implementation of regional nitrogen management policies. Results highlight a persistently negative impact of agricultural land use, the importance of sustained policies and long-term monitoring to mitigate nitrogen legacy effects, and the need for improved datasets to better capture controlling factors.
Haegyeong Lee, Manuel Gossler, Kai Zosseder, Philipp Blum, Peter Bayer, and Gabriel C. Rau
Hydrol. Earth Syst. Sci., 29, 1359–1378, https://doi.org/10.5194/hess-29-1359-2025, https://doi.org/10.5194/hess-29-1359-2025, 2025
Short summary
Short summary
A systematic laboratory experiment elucidates two-phase heat transport due to water flow in saturated porous media to understand thermal propagation in aquifers. Results reveal delayed thermal arrival in the solid phase, depending on grain size and flow velocity. Analytical modeling using standard local thermal equilibrium (LTE) and advanced local thermal non-equilibrium (LTNE) theory fails to describe temperature breakthrough curves, highlighting the need for more advanced numerical approaches.
Vianney Sivelle, Guillaume Cinkus, Naomi Mazzilli, David Labat, Bruno Arfib, Nicolas Massei, Yohann Cousquer, Dominique Bertin, and Hervé Jourde
Hydrol. Earth Syst. Sci., 29, 1259–1276, https://doi.org/10.5194/hess-29-1259-2025, https://doi.org/10.5194/hess-29-1259-2025, 2025
Short summary
Short summary
KarstMod provides a platform for global modelling of the rain level–flow relationship in karstic basins. The platform provides a set of tools to assess the dynamics of the compartments considered in the model and to detect possible flaws in structure and parameterization. This platform is developed as part of the French observatory network on karst hydrology (SNO KARST), which aims to strengthen the sharing of knowledge and promote interdisciplinary research on karst systems at a national level.
Attila Galsa, Márk Szijártó, Ádám Tóth, and Judit Mádl-Szőnyi
EGUsphere, https://doi.org/10.5194/egusphere-2025-559, https://doi.org/10.5194/egusphere-2025-559, 2025
Short summary
Short summary
Understanding groundwater flow is crucial for both environmental and economic reasons. In our study, the dynamic interaction of the forces driving groundwater flow is presented, partly in synthetic models and partly in a real deep geothermal reservoir. We point out that ignoring certain driving forces can lead to oversimplification of groundwater flow and even misinterpretation of the phenomena, causing environmental problems or economic losses.
Sivarama Krishna Reddy Chidepudi, Nicolas Massei, Abderrahim Jardani, Bastien Dieppois, Abel Henriot, and Matthieu Fournier
Hydrol. Earth Syst. Sci., 29, 841–861, https://doi.org/10.5194/hess-29-841-2025, https://doi.org/10.5194/hess-29-841-2025, 2025
Short summary
Short summary
This study explores how deep learning can improve our understanding of groundwater levels, using an approach that combines climate data and physical characteristics of aquifers. By focusing on different types of groundwater levels and employing techniques like clustering and wavelet transform, the study highlights the importance of targeting relevant information. This research not only advances groundwater simulation but also emphasizes the benefits of different modelling approaches.
Han Cao, Jinlong Qian, Huanliang Chen, Chunwei Liu, Minghui Lyu, Shuai Gao, Weihong Dong, and Caiping Hu
EGUsphere, https://doi.org/10.5194/egusphere-2025-281, https://doi.org/10.5194/egusphere-2025-281, 2025
Short summary
Short summary
This research employed multi-methods to quantitatively analyze the impacts of managed aquifer recharge (MAR) and exploitation on the dynamics of groundwater level and quality in a typical northern China karst area, the Baotu Spring area in Jinan City. The complementary techniques enhance the accuracy of quantitative research. The results indicate that the MAR and exploitation in Baotu Spring area significantly impact karst groundwater level and quality.
Igor Karlović, Mitja Janža, Edmundo Placencia-Gómez, and Tamara Marković
EGUsphere, https://doi.org/10.5194/egusphere-2025-327, https://doi.org/10.5194/egusphere-2025-327, 2025
Short summary
Short summary
This study presents a methodology for understanding heterogeneity in alluvial aquifers by combining borehole data, electrical resistivity tomography, and stochastic modeling. The approach uses hydrofacies to incorporate spatial variability into groundwater models. The hydrofacies distribution helps define hydraulic conductivity fields, essential for groundwater flow and contaminant transport simulations, providing a valuable tool for sustainable aquifer management in sedimentary environments.
Friederike Currle, René Therrien, and Oliver S. Schilling
EGUsphere, https://doi.org/10.5194/egusphere-2025-372, https://doi.org/10.5194/egusphere-2025-372, 2025
Short summary
Short summary
We present a new approach to simulate the transport of microbes in river-aquifer systems in the integrated hydrological model HydroGeoSphere. Compared to existing models, the advantage of the new implementation lies in the consideration of all relevant parts of the water budget and the flexibility to simulate in parallel the reactive transport of several microbial species and solutes. The new developed tool enables to improve our understanding of pathogen transport in river-groundwater systems.
Aohan Jin, Wenguang Shi, Renjie Zhou, Hongbin Zhan, Quanrong Wang, and Xuan Gu
EGUsphere, https://doi.org/10.5194/egusphere-2024-4145, https://doi.org/10.5194/egusphere-2024-4145, 2025
Short summary
Short summary
This study developed a novel deep transfer learning (DTL) approach, which integrates the physical mechanisms from an analytical model using a transfer learning technique. Results indicate that the DTL model maintains satisfactory performance even in heterogeneous conditions, with uncertainties in observations and sparse training data compared to the DNN model.
Etienne Marti, Sarah Leray, and Clément Roques
Hydrol. Earth Syst. Sci. Discuss., https://doi.org/10.5194/hess-2024-381, https://doi.org/10.5194/hess-2024-381, 2024
Revised manuscript accepted for HESS
Short summary
Short summary
We show that the response of groundwater-dependent wetlands to recharge changes can be predicted based on landform properties, providing a practical approach for wetland vulnerability assessment. We reveal that mountain catchments are less sensitive to recharge changes than lowland catchments. It offers insights for evaluating the vulnerability of catchments to climate change impacts and has direct implications for water resource management and conservation planning in diverse landscapes.
Raoul A. Collenteur, Ezra Haaf, Mark Bakker, Tanja Liesch, Andreas Wunsch, Jenny Soonthornrangsan, Jeremy White, Nick Martin, Rui Hugman, Ed de Sousa, Didier Vanden Berghe, Xinyang Fan, Tim J. Peterson, Jānis Bikše, Antoine Di Ciacca, Xinyue Wang, Yang Zheng, Maximilian Nölscher, Julian Koch, Raphael Schneider, Nikolas Benavides Höglund, Sivarama Krishna Reddy Chidepudi, Abel Henriot, Nicolas Massei, Abderrahim Jardani, Max Gustav Rudolph, Amir Rouhani, J. Jaime Gómez-Hernández, Seifeddine Jomaa, Anna Pölz, Tim Franken, Morteza Behbooei, Jimmy Lin, and Rojin Meysami
Hydrol. Earth Syst. Sci., 28, 5193–5208, https://doi.org/10.5194/hess-28-5193-2024, https://doi.org/10.5194/hess-28-5193-2024, 2024
Short summary
Short summary
We show the results of the 2022 Groundwater Time Series Modelling Challenge; 15 teams applied data-driven models to simulate hydraulic heads, and three model groups were identified: lumped, machine learning, and deep learning. For all wells, reasonable performance was obtained by at least one team from each group. There was not one team that performed best for all wells. In conclusion, the challenge was a successful initiative to compare different models and learn from each other.
Steven Reinaldo Rusli, Victor F. Bense, Syed M. T. Mustafa, and Albrecht H. Weerts
Hydrol. Earth Syst. Sci., 28, 5107–5131, https://doi.org/10.5194/hess-28-5107-2024, https://doi.org/10.5194/hess-28-5107-2024, 2024
Short summary
Short summary
In this paper, we investigate the impact of climatic and anthropogenic factors on future groundwater availability. The changes are simulated using hydrological and groundwater flow models. We find that future groundwater status is influenced more by anthropogenic factors than climatic factors. The results are beneficial for informing responsible parties in operational water management about achieving future (ground)water governance.
Mariana Gomez, Maximilian Nölscher, Andreas Hartmann, and Stefan Broda
Hydrol. Earth Syst. Sci., 28, 4407–4425, https://doi.org/10.5194/hess-28-4407-2024, https://doi.org/10.5194/hess-28-4407-2024, 2024
Short summary
Short summary
To understand the impact of external factors on groundwater level modelling using a 1-D convolutional neural network (CNN) model, we train, validate, and tune individual CNN models for 505 wells distributed across Lower Saxony, Germany. We then evaluate the performance of these models against available geospatial and time series features. This study provides new insights into the relationship between these factors and the accuracy of groundwater modelling.
Nariman Mahmoodi, Ulrich Struck, Michael Schneider, and Christoph Merz
Hydrol. Earth Syst. Sci. Discuss., https://doi.org/10.5194/hess-2024-214, https://doi.org/10.5194/hess-2024-214, 2024
Revised manuscript accepted for HESS
Short summary
Short summary
Understanding water balance in lakes is complex. We studied Lake Gross Glienicke in Germany, using an innovative method that combines isotope measurements and a hydrological model to improve estimates of water inflow and evaporation. Our findings show a high correlation between the two approaches, leading to better predictions of lake water dynamics. This research offers a reliable way to evaluate the model outputs.
Laurent Gourdol, Michael K. Stewart, Uwe Morgenstern, and Laurent Pfister
Hydrol. Earth Syst. Sci., 28, 3519–3547, https://doi.org/10.5194/hess-28-3519-2024, https://doi.org/10.5194/hess-28-3519-2024, 2024
Short summary
Short summary
Determining water transit times in aquifers is key to a better understanding of groundwater resources and their sustainable management. For our research, we used high-accuracy tritium data from 35 springs draining the Luxembourg Sandstone aquifer. We assessed the mean transit times of groundwater and found that water moves on average more than 10 times more slowly vertically in the vadose zone of the aquifer (~12 m yr-1) than horizontally in its saturated zone (~170 m yr-1).
Anna Pazola, Mohammad Shamsudduha, Jon French, Alan M. MacDonald, Tamiru Abiye, Ibrahim Baba Goni, and Richard G. Taylor
Hydrol. Earth Syst. Sci., 28, 2949–2967, https://doi.org/10.5194/hess-28-2949-2024, https://doi.org/10.5194/hess-28-2949-2024, 2024
Short summary
Short summary
This study advances groundwater research using a high-resolution random forest model, revealing new recharge areas and spatial variability, mainly in humid regions. Limited data in rainy zones is a constraint for the model. Our findings underscore the promise of machine learning for large-scale groundwater modelling while further emphasizing the importance of data collection for robust results.
Andreas Wunsch, Tanja Liesch, and Nico Goldscheider
Hydrol. Earth Syst. Sci., 28, 2167–2178, https://doi.org/10.5194/hess-28-2167-2024, https://doi.org/10.5194/hess-28-2167-2024, 2024
Short summary
Short summary
Seasons have a strong influence on groundwater levels, but relationships are complex and partly unknown. Using data from wells in Germany and an explainable machine learning approach, we showed that summer precipitation is the key factor that controls the severeness of a low-water period in fall; high summer temperatures do not per se cause stronger decreases. Preceding winters have only a minor influence on such low-water periods in general.
Evgeny Shavelzon and Yaniv Edery
Hydrol. Earth Syst. Sci., 28, 1803–1826, https://doi.org/10.5194/hess-28-1803-2024, https://doi.org/10.5194/hess-28-1803-2024, 2024
Short summary
Short summary
We investigate the interaction of transport with dissolution–precipitation reactions in porous media using the concepts of entropy and work to quantify the emergence of preferential flow paths. We show that the preferential-flow-path phenomenon and the hydraulic power required to maintain the driving pressure drop intensify over time along with the heterogeneity due to the interaction between the transport and the reactive processes. This is more pronounced in diffusion-dominated flows.
Stephen Lee, Dylan J. Irvine, Clément Duvert, Gabriel C. Rau, and Ian Cartwright
Hydrol. Earth Syst. Sci., 28, 1771–1790, https://doi.org/10.5194/hess-28-1771-2024, https://doi.org/10.5194/hess-28-1771-2024, 2024
Short summary
Short summary
Global groundwater recharge studies collate recharge values estimated using different methods that apply to different timescales. We develop a recharge prediction model, based solely on chloride, to produce a recharge map for Australia. We reveal that climate and vegetation have the most significant influence on recharge variability in Australia. Our recharge rates were lower than other models due to the long timescale of chloride in groundwater. Our method can similarly be applied globally.
Yiming Li, Uwe Schneidewind, Zhang Wen, Stefan Krause, and Hui Liu
Hydrol. Earth Syst. Sci., 28, 1751–1769, https://doi.org/10.5194/hess-28-1751-2024, https://doi.org/10.5194/hess-28-1751-2024, 2024
Short summary
Short summary
Meandering rivers are an integral part of many landscapes around the world. Here we used a new modeling approach to look at how the slope of riverbanks influences water flow and solute transport from a meandering river channel through its bank and into/out of the connected groundwater compartment (aquifer). We found that the bank slope can be a significant factor to be considered, especially when bank slope angles are small, and riverbank and aquifer conditions only allow for slow water flow.
Yiqun Gan and Quanrong Wang
Hydrol. Earth Syst. Sci., 28, 1317–1323, https://doi.org/10.5194/hess-28-1317-2024, https://doi.org/10.5194/hess-28-1317-2024, 2024
Short summary
Short summary
1. A revised 3D model of solute transport is developed in the well–aquifer system. 2. The accuracy of the new model is tested against benchmark analytical solutions. 3. Previous models overestimate the concentration of solute in both aquifers and wellbores in the injection well test case. 4. Previous models underestimate the concentration in the extraction well test case.
Annika Nolte, Ezra Haaf, Benedikt Heudorfer, Steffen Bender, and Jens Hartmann
Hydrol. Earth Syst. Sci., 28, 1215–1249, https://doi.org/10.5194/hess-28-1215-2024, https://doi.org/10.5194/hess-28-1215-2024, 2024
Short summary
Short summary
This study examines about 8000 groundwater level (GWL) time series from five continents to explore similarities in groundwater systems at different scales. Statistical metrics and machine learning techniques are applied to identify common GWL dynamics patterns and analyze their controlling factors. The study also highlights the potential and limitations of this data-driven approach to improve our understanding of groundwater recharge and discharge processes.
Tom Müller, Matteo Roncoroni, Davide Mancini, Stuart N. Lane, and Bettina Schaefli
Hydrol. Earth Syst. Sci., 28, 735–759, https://doi.org/10.5194/hess-28-735-2024, https://doi.org/10.5194/hess-28-735-2024, 2024
Short summary
Short summary
We investigate the role of a newly formed floodplain in an alpine glaciated catchment to store and release water. Based on field measurements, we built a numerical model to simulate the water fluxes and show that recharge occurs mainly due to the ice-melt-fed river. We identify three future floodplains, which could emerge from glacier retreat, and show that their combined storage leads to some additional groundwater storage but contributes little additional baseflow for the downstream river.
Benedikt Heudorfer, Tanja Liesch, and Stefan Broda
Hydrol. Earth Syst. Sci., 28, 525–543, https://doi.org/10.5194/hess-28-525-2024, https://doi.org/10.5194/hess-28-525-2024, 2024
Short summary
Short summary
We build a neural network to predict groundwater levels from monitoring wells. We predict all wells at the same time, by learning the differences between wells with static features, making it an entity-aware global model. This works, but we also test different static features and find that the model does not use them to learn exactly how the wells are different, but only to uniquely identify them. As this model class is not actually entity aware, we suggest further steps to make it so.
Trine Enemark, Rasmus Bødker Madsen, Torben O. Sonnenborg, Lærke Therese Andersen, Peter B. E. Sandersen, Jacob Kidmose, Ingelise Møller, Thomas Mejer Hansen, Karsten Høgh Jensen, and Anne-Sophie Høyer
Hydrol. Earth Syst. Sci., 28, 505–523, https://doi.org/10.5194/hess-28-505-2024, https://doi.org/10.5194/hess-28-505-2024, 2024
Short summary
Short summary
In this study, we demonstrate an approach to evaluate the interpretation uncertainty within a manually interpreted geological model in a groundwater model. Using qualitative estimates of uncertainties, several geological realizations are developed and implemented in groundwater models. We confirm existing evidence that if the conceptual model is well defined, interpretation uncertainties within the conceptual model have limited impact on groundwater model predictions.
Yong Zhang, Graham E. Fogg, HongGuang Sun, Donald M. Reeves, Roseanna M. Neupauer, and Wei Wei
Hydrol. Earth Syst. Sci., 28, 179–203, https://doi.org/10.5194/hess-28-179-2024, https://doi.org/10.5194/hess-28-179-2024, 2024
Short summary
Short summary
Pollutant release history and source identification are helpful for managing water resources, but it remains a challenge to reliably identify such information for real-world, complex transport processes in rivers and aquifers. In this study, we filled this knowledge gap by deriving a general backward governing equation and developing the efficient solver. Field applications showed that this model and solver are applicable for a broad range of flow systems, dimensions, and spatiotemporal scales.
Mennatullah T. Elrashidy, Andrew M. Ireson, and Saman Razavi
Hydrol. Earth Syst. Sci., 27, 4595–4608, https://doi.org/10.5194/hess-27-4595-2023, https://doi.org/10.5194/hess-27-4595-2023, 2023
Short summary
Short summary
Wetlands are important ecosystems that store carbon and play a vital role in the water cycle. However, hydrological computer models do not always represent wetlands and their interaction with groundwater accurately. We tested different possible ways to include groundwater–wetland interactions in these models. We found that the optimal method to include wetlands and groundwater in the models is reliant on the intended use of the models and the characteristics of the land and soil being studied.
Conny Tschritter, Christopher J. Daughney, Sapthala Karalliyadda, Brioch Hemmings, Uwe Morgenstern, and Catherine Moore
Hydrol. Earth Syst. Sci., 27, 4295–4316, https://doi.org/10.5194/hess-27-4295-2023, https://doi.org/10.5194/hess-27-4295-2023, 2023
Short summary
Short summary
Understanding groundwater travel time (groundwater age) is crucial for tracking flow and contaminants. While groundwater age is usually inferred from age tracers, this study utilised two machine learning techniques with common groundwater chemistry data. The results of both methods correspond to traditional approaches. They are useful where hydrochemistry data exist but age tracer data are limited. These methods could help enhance our knowledge, aiding in sustainable freshwater management.
Jose M. Bastias Espejo, Chris Turnadge, Russell S. Crosbie, Philipp Blum, and Gabriel C. Rau
Hydrol. Earth Syst. Sci., 27, 3447–3462, https://doi.org/10.5194/hess-27-3447-2023, https://doi.org/10.5194/hess-27-3447-2023, 2023
Short summary
Short summary
Analytical models estimate subsurface properties from subsurface–tidal load interactions. However, they have limited accuracy in representing subsurface physics and parameter estimation. We derived a new analytical solution which models flow to wells due to atmospheric tides. We applied it to field data and compared our findings with subsurface knowledge. Our results enhance understanding of subsurface systems, providing valuable information on their behavior.
Ronan Abhervé, Clément Roques, Alexandre Gauvain, Laurent Longuevergne, Stéphane Louaisil, Luc Aquilina, and Jean-Raynald de Dreuzy
Hydrol. Earth Syst. Sci., 27, 3221–3239, https://doi.org/10.5194/hess-27-3221-2023, https://doi.org/10.5194/hess-27-3221-2023, 2023
Short summary
Short summary
We propose a model calibration method constraining groundwater seepage in the hydrographic network. The method assesses the hydraulic properties of aquifers in regions where perennial streams are directly fed by groundwater. The estimated hydraulic conductivity appear to be highly sensitive to the spatial extent and density of streams. Such an approach improving subsurface characterization from surface information is particularly interesting for ungauged basins.
Amanda Triplett and Laura E. Condon
Hydrol. Earth Syst. Sci., 27, 2763–2785, https://doi.org/10.5194/hess-27-2763-2023, https://doi.org/10.5194/hess-27-2763-2023, 2023
Short summary
Short summary
Accelerated melting in mountains is a global phenomenon. The Heihe River basin depends on upstream mountains for its water supply. We built a hydrologic model to examine how shifts in streamflow and warming will impact ground and surface water interactions. The results indicate that degrading permafrost has a larger effect than melting glaciers. Additionally, warming temperatures tend to have more impact than changes to streamflow. These results can inform other mountain–valley system studies.
Guillaume Cinkus, Andreas Wunsch, Naomi Mazzilli, Tanja Liesch, Zhao Chen, Nataša Ravbar, Joanna Doummar, Jaime Fernández-Ortega, Juan Antonio Barberá, Bartolomé Andreo, Nico Goldscheider, and Hervé Jourde
Hydrol. Earth Syst. Sci., 27, 1961–1985, https://doi.org/10.5194/hess-27-1961-2023, https://doi.org/10.5194/hess-27-1961-2023, 2023
Short summary
Short summary
Numerous modelling approaches can be used for studying karst water resources, which can make it difficult for a stakeholder or researcher to choose the appropriate method. We conduct a comparison of two widely used karst modelling approaches: artificial neural networks (ANNs) and reservoir models. Results show that ANN models are very flexible and seem great for reproducing high flows. Reservoir models can work with relatively short time series and seem to accurately reproduce low flows.
Wenguang Shi, Quanrong Wang, Hongbin Zhan, Renjie Zhou, and Haitao Yan
Hydrol. Earth Syst. Sci., 27, 1891–1908, https://doi.org/10.5194/hess-27-1891-2023, https://doi.org/10.5194/hess-27-1891-2023, 2023
Short summary
Short summary
The mechanism of radial dispersion is important for understanding reactive transport in the subsurface and for estimating aquifer parameters required in the optimization design of remediation strategies. A general model and associated analytical solutions are developed in this study. The new model represents the most recent advancement on radial dispersion studies and incorporates a host of important processes that are not taken into consideration in previous investigations.
Vanja Travaš, Luka Zaharija, Davor Stipanić, and Siniša Družeta
Hydrol. Earth Syst. Sci., 27, 1343–1359, https://doi.org/10.5194/hess-27-1343-2023, https://doi.org/10.5194/hess-27-1343-2023, 2023
Short summary
Short summary
In order to model groundwater flow in karst aquifers, it is necessary to approximate the influence of the unknown and irregular structure of the karst conduits. For this purpose, a procedure based on inverse modeling is adopted. Moreover, in order to reconstruct the functional dependencies related to groundwater flow, the particle swarm method was used, through which the optimal solution of unknown functions is found by imitating the movement of ants in search of food.
Shouchuan Zhang, Zheming Shi, Guangcai Wang, Zuochen Zhang, and Huaming Guo
Hydrol. Earth Syst. Sci., 27, 401–415, https://doi.org/10.5194/hess-27-401-2023, https://doi.org/10.5194/hess-27-401-2023, 2023
Short summary
Short summary
We documented the step-like increases of water level, flow rate, and water temperatures in a confined aquifer following multiple earthquakes. By employing tidal analysis and a coupled temperature and flow rate model, we find that post-seismic vertical permeability changes and recharge model could explain the co-seismic response. And co-seismic temperature changes are caused by mixing of different volumes of water, with the mixing ratio varying according to each earthquake.
Xiaoying Zhang, Fan Dong, Guangquan Chen, and Zhenxue Dai
Hydrol. Earth Syst. Sci., 27, 83–96, https://doi.org/10.5194/hess-27-83-2023, https://doi.org/10.5194/hess-27-83-2023, 2023
Short summary
Short summary
In a data-driven framework, groundwater levels can generally only be calculated 1 time step ahead. We discuss the advance prediction with longer forecast periods rather than single time steps by constructing a model based on a temporal convolutional network. Model accuracy and efficiency were further compared with an LSTM-based model. The two models derived in this study can help people cope with the uncertainty of what might occur in hydrological scenarios under the threat of climate change.
Dimitri Rambourg, Raphaël Di Chiara, and Philippe Ackerer
Hydrol. Earth Syst. Sci., 26, 6147–6162, https://doi.org/10.5194/hess-26-6147-2022, https://doi.org/10.5194/hess-26-6147-2022, 2022
Short summary
Short summary
The reproduction of flows and contaminations underground requires a good estimation of the parameters of the geological environment (mainly permeability and porosity), in three dimensions. While most researchers rely on geophysical methods, which are costly and difficult to implement in the field, this study proposes an alternative using data that are already widely available: piezometric records (monitoring of the water table) and the lithological description of the piezometric wells.
Raphael Schneider, Julian Koch, Lars Troldborg, Hans Jørgen Henriksen, and Simon Stisen
Hydrol. Earth Syst. Sci., 26, 5859–5877, https://doi.org/10.5194/hess-26-5859-2022, https://doi.org/10.5194/hess-26-5859-2022, 2022
Short summary
Short summary
Hydrological models at high spatial resolution are computationally expensive. However, outputs from such models, such as the depth of the groundwater table, are often desired in high resolution. We developed a downscaling algorithm based on machine learning that allows us to increase spatial resolution of hydrological model outputs, alleviating computational burden. We successfully applied the downscaling algorithm to the climate-change-induced impacts on the groundwater table across Denmark.
Luca Guillaumot, Laurent Longuevergne, Jean Marçais, Nicolas Lavenant, and Olivier Bour
Hydrol. Earth Syst. Sci., 26, 5697–5720, https://doi.org/10.5194/hess-26-5697-2022, https://doi.org/10.5194/hess-26-5697-2022, 2022
Short summary
Short summary
Recharge, defining the renewal rate of groundwater resources, is difficult to estimate at basin scale. Here, recharge variations are inferred from water table variations recorded in boreholes. First, results show that aquifer-scale properties controlling these variations can be inferred from boreholes. Second, groundwater is recharged by both intense and seasonal rainfall. Third, the short-term contribution appears overestimated in recharge models and depends on the unsaturated zone thickness.
Alberto Casillas-Trasvina, Bart Rogiers, Koen Beerten, Laurent Wouters, and Kristine Walraevens
Hydrol. Earth Syst. Sci., 26, 5577–5604, https://doi.org/10.5194/hess-26-5577-2022, https://doi.org/10.5194/hess-26-5577-2022, 2022
Short summary
Short summary
Heat in the subsurface can be used to characterize aquifer flow behaviour. The temperature data obtained can be useful for understanding the groundwater flow, which is of particular importance in waste disposal studies. Satellite images of surface temperature and a temperature–time curve were implemented in a heat transport model. Results indicate that conduction plays a major role in the aquifer and support the usefulness of temperature measurements.
Tunde Olarinoye, Tom Gleeson, and Andreas Hartmann
Hydrol. Earth Syst. Sci., 26, 5431–5447, https://doi.org/10.5194/hess-26-5431-2022, https://doi.org/10.5194/hess-26-5431-2022, 2022
Short summary
Short summary
Analysis of karst spring recession is essential for management of groundwater. In karst, recession is dominated by slow and fast components; separating these components is by manual and subjective approaches. In our study, we tested the applicability of automated streamflow recession extraction procedures for a karst spring. Results showed that, by simple modification, streamflow extraction methods can identify slow and fast components: derived recession parameters are within reasonable ranges.
Min Lu, Bart Rogiers, Koen Beerten, Matej Gedeon, and Marijke Huysmans
Hydrol. Earth Syst. Sci., 26, 3629–3649, https://doi.org/10.5194/hess-26-3629-2022, https://doi.org/10.5194/hess-26-3629-2022, 2022
Short summary
Short summary
Lowland rivers and shallow aquifers are closely coupled. We study their interactions here using a combination of impulse response modeling and hydrological data analysis. The results show that the lowland catchments are groundwater dominated and that the hydrological system from precipitation impulse to groundwater inflow response is a very fast response regime. This study also provides an alternative method to estimate groundwater inflow to rivers from the perspective of groundwater level.
Zhongxia Li, Junwei Wan, Tao Xiong, Hongbin Zhan, Linqing He, and Kun Huang
Hydrol. Earth Syst. Sci., 26, 3359–3375, https://doi.org/10.5194/hess-26-3359-2022, https://doi.org/10.5194/hess-26-3359-2022, 2022
Short summary
Short summary
Four permeable rocks with different pore sizes were considered to provide experimental evidence of Forchheimer flow and the transition between different flow regimes. The mercury injection technique was used to measure the pore size distribution, which is an essential factor for determining the flow regime, for four permeable stones. Finally, the influences of porosity and particle size on the Forchheimer coefficients were discussed.
Andreas Wunsch, Tanja Liesch, Guillaume Cinkus, Nataša Ravbar, Zhao Chen, Naomi Mazzilli, Hervé Jourde, and Nico Goldscheider
Hydrol. Earth Syst. Sci., 26, 2405–2430, https://doi.org/10.5194/hess-26-2405-2022, https://doi.org/10.5194/hess-26-2405-2022, 2022
Short summary
Short summary
Modeling complex karst water resources is difficult enough, but often there are no or too few climate stations available within or close to the catchment to deliver input data for modeling purposes. We apply image recognition algorithms to time-distributed, spatially gridded meteorological data to simulate karst spring discharge. Our models can also learn the approximate catchment location of a spring independently.
Brian Berkowitz
Hydrol. Earth Syst. Sci., 26, 2161–2180, https://doi.org/10.5194/hess-26-2161-2022, https://doi.org/10.5194/hess-26-2161-2022, 2022
Short summary
Short summary
Extensive efforts have focused on quantifying conservative chemical transport in geological formations. We assert that an explicit accounting of temporal information, under uncertainty, in addition to spatial information, is fundamental to an effective modeling formulation. We further assert that efforts to apply chemical transport equations at large length scales, based on measurements and model parameter values relevant to significantly smaller length scales, are an unattainable
holy grail.
Guilherme E. H. Nogueira, Christian Schmidt, Daniel Partington, Philip Brunner, and Jan H. Fleckenstein
Hydrol. Earth Syst. Sci., 26, 1883–1905, https://doi.org/10.5194/hess-26-1883-2022, https://doi.org/10.5194/hess-26-1883-2022, 2022
Short summary
Short summary
In near-stream aquifers, mixing between stream water and ambient groundwater can lead to dilution and the removal of substances that can be harmful to the water ecosystem at high concentrations. We used a numerical model to track the spatiotemporal evolution of different water sources and their mixing around a stream, which are rather difficult in the field. Results show that mixing mainly develops as narrow spots, varying In time and space, and is affected by magnitudes of discharge events.
Jacques Bodin, Gilles Porel, Benoît Nauleau, and Denis Paquet
Hydrol. Earth Syst. Sci., 26, 1713–1726, https://doi.org/10.5194/hess-26-1713-2022, https://doi.org/10.5194/hess-26-1713-2022, 2022
Short summary
Short summary
Assessment of the karst network geometry is an important challenge in the accurate modeling of karst aquifers. In this study, we propose an approach for the identification of effective three-dimensional discrete karst conduit networks conditioned on tracer tests and geophysical data. The applicability of the proposed approach is illustrated through a case study at the Hydrogeological Experimental Site in Poitiers, France.
Zexuan Xu, Rebecca Serata, Haruko Wainwright, Miles Denham, Sergi Molins, Hansell Gonzalez-Raymat, Konstantin Lipnikov, J. David Moulton, and Carol Eddy-Dilek
Hydrol. Earth Syst. Sci., 26, 755–773, https://doi.org/10.5194/hess-26-755-2022, https://doi.org/10.5194/hess-26-755-2022, 2022
Short summary
Short summary
Climate change could change the groundwater system and threaten water supply. To quantitatively evaluate its impact on water quality, numerical simulations with chemical and reaction processes are required. With the climate projection dataset, we used the newly developed hydrological and chemical model to investigate the movement of contaminants and assist the management of contamination sites.
Cited articles
Addimando, N., Engel, M., Schwarz, F., and Batič, M.: A DEEP LEARNING APPROACH FOR CROP TYPE MAPPING BASED ON COMBINED TIME SERIES OF SATELLITE AND WEATHER DATA, The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLIII-B3-2022, 1301–1308, https://doi.org/10.5194/isprs-archives-XLIII-B3-2022-1301-2022, 2022. a
Althoff, D., Rodrigues, L. N., and Bazame, H. C.: Uncertainty quantification for hydrological models based on neural networks: the dropout ensemble, Stoch. Env. Res. Risk A., 35, 1051–1067, https://doi.org/10.1007/s00477-021-01980-8, 2021. a
Amatulli, G., Domisch, S., Tuanmu, M.-N., Parmentier, B., Ranipeta, A., Malczyk, J., and Jetz, W.: A suite of global, cross-scale topographic variables for environmental and biodiversity modeling, Sci. Data, 5, 180040, https://doi.org/10.1038/sdata.2018.40, 2018. a
Bakalowicz, M.: Karst groundwater: a challenge for new resources, Hydrogeol. J., 13, 148–160, https://doi.org/10.1007/s10040-004-0402-9, 2005. a
Beven, K. J. and Kirkby, M. J.: A physically based, variable contributing area model of basin hydrology / Un modèle à base physique de zone d'appel variable de l'hydrologie du bassin versant, Hydrol. Sci. B., 24, 43–69, https://doi.org/10.1080/02626667909491834, 1979. a, b
BGR: Bodenarten in Oberböden Deutschlands 1:1.000.000 (BOART1000), Digitaler Datenbestand, version 2.0, UUID: 410B839B-DCF2-45BE-8788-3C65BBEC2D59, https://geoportal.bgr.de/mapapps/resources/apps/geoportal/index.html?lang=de#/datasets/portal/DADB8BB6-4A7A-4CB7-908D-EA0767B068D7 (last access: 14 October 2024), 2007. a
BGR: Mittlere jährliche Grundwasserneubildung von Deutschland 1:1.000.000 (GWN1000), Digital map data, version 1, https://geoportal.bgr.de/mapapps/resources/apps/geoportal/index.html?lang=de#/datasets/portal/40E14FF1-99D4-43DA-AF7B-C039F0463BF8 (last access: 14 October 2024), 2019. a
BGR and SGD: Hydrogeologische Übersichtskarte von Deutschland 1:250.000 (HÃœK250), Digitaler Datenbestand, version 1.0.3, UUID: afbc7573-d18d-453e-bff3-6477af048770, https://geoportal.bgr.de/mapapps/resources/apps/geoportal/index.html?lang=de#/datasets/portal/61ac4628-6b62-48c6-89b8-46270819f0d6 (last access: 14 October 2024), 2019. a, b, c, d
Buchhorn, M., Smets, B., Bertels, L., De Roo, B., Lesiv, M., Tsendbazar, N. E., Linlin, L., and Tarko, A.: opernicus Global Land Service: Land Cover 100m: Version 3 Globe 2015–2019: Product User Manual, Zenodo, https://doi.org/10.5281/zenodo.3938963, 2020. a
Cao, D., Wang, Y., Duan, J., Zhang, C., Zhu, X., Huang, C., Tong, Y., Xu, B., Bai, J., Tong, J., and Zhang, Q.: Spectral Temporal Graph Neural Network for Multivariate Time-series Forecasting, arXiv [preprint], https://doi.org/10.48550/arXiv.2103.07719, 2021. a
Challu, C., Olivares, K. G., Oreshkin, B. N., Garza, F., Mergenthaler-Canseco, M., and Dubrawski, A.: N-HiTS: Neural Hierarchical Interpolation for Time Series Forecasting, arXiv [preprint], https://doi.org/10.48550/arXiv.2201.12886, 2022. a, b, c
Chiaburu, T. and Bießmann, F.: Interpretable Time Series Models for Wastewater Modeling in Combined Sewer Overflows, Proced. Comput. Sci., 237, 155–162, https://doi.org/10.1016/j.procs.2024.05.091, 2024. a
Chidepudi, S. K. R., Massei, N., Jardani, A., and Henriot, A.: Groundwater level reconstruction using long-term climate reanalysis data and deep neural networks, J. Hydrol.-Regional Studies, 51, 101632, https://doi.org/10.1016/j.ejrh.2023.101632, 2024. a
Chidepudi, S. K. R., Massei, N., Jardani, A., Dieppois, B., Henriot, A., and Fournier, M.: Training deep learning models with a multi-station approach and static aquifer attributes for groundwater level simulation: what is the best way to leverage regionalised information?, Hydrol. Earth Syst. Sci., 29, 841–861, https://doi.org/10.5194/hess-29-841-2025, 2025. a
Collenteur, R. A., Haaf, E., Bakker, M., Liesch, T., Wunsch, A., Soonthornrangsan, J., White, J., Martin, N., Hugman, R., de Sousa, E., Vanden Berghe, D., Fan, X., Peterson, T. J., Bikše, J., Di Ciacca, A., Wang, X., Zheng, Y., Nölscher, M., Koch, J., Schneider, R., Benavides Höglund, N., Krishna Reddy Chidepudi, S., Henriot, A., Massei, N., Jardani, A., Rudolph, M. G., Rouhani, A., Gómez-Hernández, J. J., Jomaa, S., Pölz, A., Franken, T., Behbooei, M., Lin, J., and Meysami, R.: Data-driven modelling of hydraulic-head time series: results and lessons learned from the 2022 Groundwater Time Series Modelling Challenge, Hydrol. Earth Syst. Sci., 28, 5193–5208, https://doi.org/10.5194/hess-28-5193-2024, 2024. a
Das, A., Kong, W., Leach, A., Mathur, S., Sen, R., and Yu, R.: Long-term Forecasting with TiDE: Time-series Dense Encoder, arXiv [preprint], https://doi.org/10.48550/arXiv.2304.08424, 2024. a
European Environment Agency: European Digital Elevation Model (EU-DEM) v1.1, European Environment Agency (EEA) [code], 2018. a
Gneiting, T. and Raftery, A. E.: Strictly proper scoring rules, prediction, and estimation, J. Am. Stat. Assoc., 102, 359–378, https://doi.org/10.1198/016214506000001437, 2007. a
Gomez, M., Nölscher, M., Hartmann, A., and Broda, S.: Assessing groundwater level modelling using a 1-D convolutional neural network (CNN): linking model performances to geospatial and time series features, Hydrol. Earth Syst. Sci., 28, 4407–4425, https://doi.org/10.5194/hess-28-4407-2024, 2024. a
Guzman, S. M., Paz, J. O., and Tagert, M. L. M.: The Use of NARX Neural Networks to Forecast Daily Groundwater Levels, Water Resour. Manag., 31, 1591–1603, https://doi.org/10.1007/s11269-017-1598-5, 2017. a
Hermans, T., Goderniaux, P., Jougnot, D., Fleckenstein, J. H., Brunner, P., Nguyen, F., Linde, N., Huisman, J. A., Bour, O., Lopez Alvis, J., Hoffmann, R., Palacios, A., Cooke, A.-K., Pardo-Álvarez, Á., Blazevic, L., Pouladi, B., Haruzi, P., Fernandez Visentini, A., Nogueira, G. E. H., Tirado-Conde, J., Looms, M. C., Kenshilikova, M., Davy, P., and Le Borgne, T.: Advancing measurements and representations of subsurface heterogeneity and dynamic processes: towards 4D hydrogeology, Hydrol. Earth Syst. Sci., 27, 255–287, https://doi.org/10.5194/hess-27-255-2023, 2023. a
Izmailov, P., Podoprikhin, D., Garipov, T., Vetrov, D., and Wilson, A. G.: Averaging Weights Leads to Wider Optima and Better Generalization, arXiv [preprint], https://doi.org/10.48550/arXiv.1803.05407, 2019. a
Jung, H., Saynisch-Wagner, J., and Schulz, S.: Can eXplainable AI Offer a New Perspective for Groundwater Recharge Estimation? – Global-Scale Modeling Using Neural Network, Water Resour. Res., 60, e2023WR036360, https://doi.org/10.1029/2023WR036360, 2024. a
Kan, K., Aubet, F.-X., Januschowski, T., Park, Y., Benidis, K., Ruthotto, L., and Gasthaus, J.: Multivariate Quantile Function Forecaster, arXiv [preprint], https://doi.org/10.48550/arXiv.2202.11316, 2022. a, b
Klotz, D., Kratzert, F., Gauch, M., Keefe Sampson, A., Brandstetter, J., Klambauer, G., Hochreiter, S., and Nearing, G.: Uncertainty estimation with deep learning for rainfall–runoff modeling, Hydrol. Earth Syst. Sci., 26, 1673–1693, https://doi.org/10.5194/hess-26-1673-2022, 2022. a
Koya, S. R. and Roy, T.: Temporal Fusion Transformers for Streamflow Prediction: Value of Combining Attention with Recurrence, arXiv [preprint], https://doi.org/10.48550/arXiv.2305.12335, 2023. a, b
Kratzert, F., Klotz, D., Shalev, G., Klambauer, G., Hochreiter, S., and Nearing, G.: Towards learning universal, regional, and local hydrological behaviors via machine learning applied to large-sample datasets, Hydrol. Earth Syst. Sci., 23, 5089–5110, https://doi.org/10.5194/hess-23-5089-2019, 2019. a
Kratzert, F., Gauch, M., Klotz, D., and Nearing, G.: HESS Opinions: Never train a Long Short-Term Memory (LSTM) network on a single basin, Hydrol. Earth Syst. Sci., 28, 4187–4201, https://doi.org/10.5194/hess-28-4187-2024, 2024. a, b
Kunz, S.: global-groundwater-models-main, GitHub [code], https://github.com/KunzstBGR/global-groundwater-models-main, last access: 14 July 2025. a
Kværner, J. and Snilsberg, P.: Groundwater hydrology of boreal peatlands above a bedrock tunnel – Drainage impacts and surface water groundwater interactions, J. Hydrol., 403, 278–291, https://doi.org/10.1016/j.jhydrol.2011.04.006, 2011. a
Li, X., Khandelwal, A., Jia, X., Cutler, K., Ghosh, R., Renganathan, A., Xu, S., Tayal, K., Nieber, J., Duffy, C., Steinbach, M., and Kumar, V.: Regionalization in a Global Hydrologic Deep Learning Model: From Physical Descriptors to Random Vectors, Water Resour. Res., 58, e2021WR031794, https://doi.org/10.1029/2021WR031794, 2022. a, b
Lim, B., Arık, S. Ö., Loeff, N., and Pfister, T.: Temporal Fusion Transformers for interpretable multi-horizon time series forecasting, Int. J. Forecasting, 37, 1748–1764, https://doi.org/10.1016/j.ijforecast.2021.03.012, 2021. a, b
Linardatos, P., Papastefanopoulos, V., Panagiotakopoulos, T., and Kotsiantis, S.: CO2 concentration forecasting in smart cities using a hybrid ARIMA–TFT model on multivariate time series IoT data, Sci. Rep., 13, 17266, https://doi.org/10.1038/s41598-023-42346-0, 2023. a
Linde, N., Ginsbourger, D., Irving, J., Nobile, F., and Doucet, A.: On uncertainty quantification in hydrogeology and hydrogeophysics, Adv. Water Resour., 110, 166–181, https://doi.org/10.1016/j.advwatres.2017.10.014, 2017. a
Makridakis, S., Spiliotis, E., and Assimakopoulos, V.: The M4 Competition: 100,000 time series and 61 forecasting methods, Int. J. Forecasting, 36, 54–74, https://doi.org/10.1016/j.ijforecast.2019.04.014, 2020. a
Makridakis, S., Spiliotis, E., and Assimakopoulos, V.: M5 accuracy competition: Results, findings, and conclusions, Int. J. Forecasting, 38, 1346–1364, https://doi.org/10.1016/j.ijforecast.2021.11.013, 2022. a
Manzoni, A., Porta, G. M., Guadagnini, L., Guadagnini, A., and Riva, M.: Probabilistic reconstruction via machine-learning of the Po watershed aquifer system (Italy), Hydrogeol. J., 31, 1547–1563, https://doi.org/10.1007/s10040-023-02677-8, 2023. a
Manzoni, A., Porta, G. M., Guadagnini, L., Guadagnini, A., and Riva, M.: A comprehensive framework for stochastic calibration and sensitivity analysis of large-scale groundwater models, Hydrol. Earth Syst. Sci., 28, 2661–2682, https://doi.org/10.5194/hess-28-2661-2024, 2024. a
Nearing, G. S., Kratzert, F., Sampson, A. K., Pelissier, C. S., Klotz, D., Frame, J. M., Prieto, C., and Gupta, H. V.: What Role Does Hydrological Science Play in the Age of Machine Learning?, Water Resour. Res., 57, e2020WR028091, https://doi.org/10.1029/2020WR028091, 2021. a
Nölscher, M., Mutz, M., and Broda, S.: Multiorder hydrologic Position for Europe – a Set of Features for Machine Learning and Analysis in Hydrology, Sci. Data, 9, 662, https://doi.org/10.1038/s41597-022-01787-4, 2022. a, b, c
Olivares, K. G., Challu, C., Marcjasz, G., Weron, R., and Dubrawski, A.: Neural basis expansion analysis with exogenous variables: Forecasting electricity prices with NBEATSx, Int. J. Forecasting, 39, 884–900, https://doi.org/10.1016/j.ijforecast.2022.03.001, 2022. a
Oosterwoud, M., van der Ploeg, M., van der Schaaf, S., and van der Zee, S.: Variation in hydrologic connectivity as a result of microtopography explained by discharge to catchment size relationship, Hydrol. Process., 31, 2683–2699, https://doi.org/10.1002/hyp.11164, 2017. a
Oreshkin, B. N., Carpov, D., Chapados, N., and Bengio, Y.: N-BEATS: Neural basis expansion analysis for interpretable time series forecasting, arXiv [preprint], https://doi.org/10.48550/arXiv.1905.10437, 2020. a
Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., Antiga, L., Desmaison, A., Köpf, A., Yang, E., DeVito, Z., Raison, M., Tejani, A., Chilamkurthy, S., Steiner, B., Fang, L., Bai, J., and Chintala, S.: PyTorch: An Imperative Style, High-Performance Deep Learning Library, arXiv [preprint], https://doi.org/10.48550/arXiv.1912.01703, 2019. a
Rangapuram, S. S., Seeger, M. W., Gasthaus, J., Stella, L., Wang, Y., and Januschowski, T.: Deep State Space Models for Time Series Forecasting. Advances in Neural Information Processing Systems, 31, https://proceedings.neurips.cc/paper/2018/file/5cf68969fb67aa6082363a6d4e6468e2-Paper.pdf (last access: 14 October 2024), 2018. a
Razafimaharo, C., Krähenmann, S., Höpp, S., Rauthe, M., and Deutschländer, T.: New high-resolution gridded dataset of daily mean, minimum, and maximum temperature and relative humidity for Central Europe (HYRAS), Theor. Appl. Climatol., 142, 1531–1553, https://doi.org/10.1007/s00704-020-03388-w, 2020. a, b, c, d
Salinas, D., Flunkert, V., Gasthaus, J., and Januschowski, T.: DeepAR: Probabilistic forecasting with autoregressive recurrent networks, Int. J. Forecasting, 36, 1181–1191, https://doi.org/10.1016/j.ijforecast.2019.07.001, 2020. a
Schroeter, L. and Gläßer, C.: Analyses and monitoring of lignite mining lakes in Eastern Germany with spectral signatures of Landsat TM satellite data, Int. J. Coal Geol., 86, 27–39, https://doi.org/10.1016/j.coal.2011.01.005, 2011. a
Sen, R., Yu, H.-F., and Dhillon, I. S.: Think Globally, Act Locally: A Deep Neural Network Approach to High-Dimensional Time Series Forecasting, arXiv [preprint], https://doi.org/10.48550/arXiv.1905.03806, 2019. a
Tao, H., Hameed, M. M., Marhoon, H. A., Zounemat-Kermani, M., Heddam, S., Kim, S., Sulaiman, S. O., Tan, M. L., Sa’adi, Z., Mehr, A. D., Allawi, M. F., Abba, S., Zain, J. M., Falah, M. W., Jamei, M., Bokde, N. D., Bayatvarkeshi, M., Al-Mukhtar, M., Bhagat, S. K., Tiyasha, T., Khedher, K. M., Al-Ansari, N., Shahid, S., and Yaseen, Z. M.: Groundwater level prediction using machine learning models: A comprehensive review, Neurocomputing, 489, 271–308, https://doi.org/10.1016/j.neucom.2022.03.014, 2022. a
Wright, L. and Demeure, N.: Ranger21: A synergistic deep learning optimizer, arXiv [preprint], https://doi.org/10.48550/arXiv.2106.13731, 2021. a
Wu, H., Zhou, H., Long, M., and Wang, J.: Interpretable weather forecasting for worldwide stations with a unified deep model, Nature Machine Intelligence, 5, 602–611, https://doi.org/10.1038/s42256-023-00667-9, 2023. a
Wunsch, A., Liesch, T., and Broda, S.: Forecasting groundwater levels using nonlinear autoregressive networks with exogenous input (NARX), J. Hydrol., 567, 743–758, https://doi.org/10.1016/j.jhydrol.2018.01.045, 2018. a
Wunsch, A., Liesch, T., and Broda, S.: Groundwater level forecasting with artificial neural networks: a comparison of long short-term memory (LSTM), convolutional neural networks (CNNs), and non-linear autoregressive networks with exogenous input (NARX), Hydrol. Earth Syst. Sci., 25, 1671–1687, https://doi.org/10.5194/hess-25-1671-2021, 2021. a, b
Wunsch, A., Liesch, T., and Broda, S.: Deep learning shows declining groundwater levels in Germany until 2100 due to climate change, Nat. Commun., 13, 1–13, https://doi.org/10.1038/s41467-022-28770-2, 2022a. a, b, c
Wunsch, A., Liesch, T., and Broda, S.: Feature-based Groundwater Hydrograph Clustering Using Unsupervised Self-Organizing Map-Ensembles, Water Resour. Manag., 36, 39–54, https://doi.org/10.1007/s11269-021-03006-y, 2022b. a, b
Short summary
Accurate groundwater level predictions are crucial for sustainable management. This study applies two machine learning models – Neural Hierarchical Interpolation for Time Series Forecasting (N-HiTS) and the Temporal Fusion Transformer (TFT) – to forecast seasonal groundwater levels for 5288 wells across Germany. N-HiTS outperformed TFT, with both models performing well in diverse hydrogeological settings, particularly in lowlands with distinct seasonal dynamics.
Accurate groundwater level predictions are crucial for sustainable management. This study...
