Articles | Volume 24, issue 12
Research article 03 Dec 2020
Research article | 03 Dec 2020
Physics-inspired integrated space–time artificial neural networks for regional groundwater flow modeling
Ali Ghaseminejad and Venkatesh Uddameri
No articles found.
Farhang Forghanparast, Elma Annette Hernandez, and Venkatesh Uddameri
Hydrol. Earth Syst. Sci. Discuss.,
Manuscript not accepted for further reviewShort summary
Intermittent flows are ecologically important yet poorly studied. New machine learning modeling methods and modeling evaluation schemes have been developed and tested over a wide range of climatic conditions to address this important knowledge gap in hydrology. The results demonstrate the superiority of these new algorithms to properly capture intermittent flow characteristics and evaluate no-flow persistence and flow regime transitions important for eco-hydrological assessments.
Related subject area
Subject: Groundwater hydrology | Techniques and Approaches: Modelling approachesCoupling saturated and unsaturated flow: comparing the iterative and the non-iterative approachTime lags of nitrate, chloride, and tritium in streams assessed by dynamic groundwater flow tracking in a lowland landscapeUsing Long Short-Term Memory networks to connect water table depth anomalies to precipitation anomalies over EuropeEstimation of groundwater recharge from groundwater levels using nonlinear transfer function noise models and comparison to lysimeter dataEarly hypogenic carbonic acid speleogenesis in unconfined limestone aquifers by upwelling deep-seated waters with high CO2 concentration: a modelling approachImpacts of climate change on groundwater flooding and ecohydrology in lowland karstHow daily groundwater table drawdown affects the diel rhythm of hyporheic exchangeGroundwater 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)Groundwater and baseflow drought responses to synthetic recharge stress testsDetermination of vadose zone and saturated zone nitrate lag times using long-term groundwater monitoring data and statistical machine learningTaking theory to the field: streamflow generation mechanisms in an intermittent, Mediterranean catchmentModelling the hydrological interactions between a fissured granite aquifer and a valley mire in the Massif Central, FranceA new criterion for determining the representative elementary volume of translucent porous media and inner contaminantHydraulic and geochemical impact of occasional saltwater intrusions through a submarine spring in a karst and thermal aquifer (Balaruc peninsula near Montpellier, France)Calibration of a lumped karst system model and application to the Qachqouch karst spring (Lebanon) under climate change conditionsSensitivity of hydrologic and geologic parameters on recharge processes in a highly heterogeneous, semi-confined aquifer systemBasin-scale multi-objective simulation-optimization modeling for conjunctive use of surface water and groundwater in northwest ChinaAssessing the response of groundwater quantity and travel time distribution to 1.5, 2, and 3 °C global warming in a mesoscale central German basinGroundwater mean residence times of a subtropical barrier sand islandOn the conceptual complexity of non-point source management: impact of spatial variabilityThe millennium-old hydrogeology textbook The Extraction of Hidden Waters by the Persian mathematician and engineer Abubakr Mohammad Karaji (953 CE–1029 CE)Modeling groundwater responses to climate change in the Prairie Pothole RegionA multi-environmental tracer study to determine groundwater residence times and recharge in a structurally complex multi-aquifer systemA three-dimensional palaeohydrogeological reconstruction of the groundwater salinity distribution in the Nile Delta AquiferModelling of the shallow water table at high spatial resolution using random forestsAn extended trajectory-mechanics approach for calculating the path of a pressure transient: travel-time tomographyGlobal sensitivity analysis and adaptive stochastic sampling of a subsurface-flow model using active subspacesA comprehensive quasi-3-D model for regional-scale unsaturated–saturated water flowDecomposition technique for contributions to groundwater heads from inside and outside of an arbitrary boundary: application to Guantao County, North China PlainHigh-resolution paleovalley classification from airborne electromagnetic imaging and deep neural network training using digital elevation model dataA partially coupled hydro-mechanical analysis of the Bengal Aquifer System under hydrological loadingReactive transport with wellbore storages in a single-well push–pull testDynamics of wormhole formation in fractured limestonesA general analytical model for head response to oscillatory pumping in unconfined aquifers: effects of delayed gravity drainage and initial conditionFaulting patterns in the Lower Yarmouk Gorge potentially influence groundwater flow pathsStochastic modeling of flow and conservative transport in three-dimensional discrete fracture networksLocality-based 3-D multiple-point statistics reconstruction using 2-D geological cross sectionsApplication of an improved global-scale groundwater model for water table estimation across New ZealandManaged aquifer recharge with reverse-osmosis desalinated seawater: modeling the spreading in groundwater using stable water isotopesNumerical modeling of flow and transport in the Bari industrial area by means of rough walled parallel plate and random walk modelsEstimation of effective porosity in large-scale groundwater models by combining particle tracking, auto-calibration and 14C datingSpatial prediction of groundwater spring potential mapping based on an adaptive neuro-fuzzy inference system and metaheuristic optimizationGroundwater origin, flow regime and geochemical evolution in arid endorheic watersheds: a case study from the Qaidam Basin, northwestern ChinaHigh-frequency NO3− isotope (δ15N, δ18O) patterns in groundwater recharge reveal that short-term changes in land use and precipitation influence nitrate contamination trendsDefining and analyzing the frequency and severity of flood events to improve risk management from a reinsurance standpointIntegrated assessment of future potential global change scenarios and their hydrological impacts in coastal aquifers – a new tool to analyse management alternatives in the Plana Oropesa-Torreblanca aquiferGroundwater withdrawal in randomly heterogeneous coastal aquifersA Bayesian approach to infer nitrogen loading rates from crop and land-use types surrounding private wells in the Central Valley, CaliforniaImpact of coastal forcing and groundwater recharge on the growth of a fresh groundwater lens in a mega-scale beach nourishmentHydrogeological effects of dredging navigable canals through lagoon shallows. A case study in Venice
Natascha Brandhorst, Daniel Erdal, and Insa Neuweiler
Hydrol. Earth Syst. Sci., 25, 4041–4059,Short summary
We compare two approaches for coupling a 2D groundwater model with multiple 1D models for the unsaturated zone. One is non-iterative and very fast. The other one is iterative and involves a new way of treating the specific yield, which is crucial for obtaining a consistent solution in both model compartments. Tested on different scenarios, this new method turns out to be slower than the non-iterative approach but more accurate and still very efficient compared to fully integrated 3D model runs.
Vince P. Kaandorp, Hans Peter Broers, Ype van der Velde, Joachim Rozemeijer, and Perry G. B. de Louw
Hydrol. Earth Syst. Sci., 25, 3691–3711,Short summary
We reconstructed historical and present-day tritium, chloride, and nitrate concentrations in stream water of a catchment using land-use-based input curves and calculated travel times of groundwater. Parameters such as the unsaturated zone thickness, mean travel time, and input patterns determine time lags between inputs and in-stream concentrations. The timescale of the breakthrough of pollutants in streams is dependent on the location of pollution in a catchment.
Yueling Ma, Carsten Montzka, Bagher Bayat, and Stefan Kollet
Hydrol. Earth Syst. Sci., 25, 3555–3575,Short summary
This study utilized spatiotemporally continuous precipitation anomaly (pra) and water table depth anomaly (wtda) data from integrated hydrologic simulation results over Europe in combination with Long Short-Term Memory (LSTM) networks to capture the time-varying and time-lagged relationship between pra and wtda in order to obtain reliable models to estimate wtda at the individual pixel level.
Raoul A. Collenteur, Mark Bakker, Gernot Klammler, and Steffen Birk
Hydrol. Earth Syst. Sci., 25, 2931–2949,Short summary
This study explores the use of nonlinear transfer function noise (TFN) models to simulate groundwater levels and estimate groundwater recharge from observed groundwater levels. A nonlinear recharge model is implemented in a TFN model to compute the recharge. The estimated recharge rates are shown to be in good agreement with the recharge observed with a lysimeter present at the case study site in Austria. The method can be used to obtain groundwater recharge rates at sub-yearly timescales.
Franci Gabrovšek and Wolfgang Dreybrodt
Hydrol. Earth Syst. Sci., 25, 2895–2913,Short summary
The evolution of karst aquifers is often governed by solutions gaining their aggressiveness in depth. Although the principles of
hypogene speleogenesisare known, modelling studies based on reactive flow in fracture networks are missing. We present a model where dissolution at depth is triggered by the mixing of waters of different origin and chemistry. We show how the initial position of the mixing zone and flow instabilities therein determine the position and shape of the final conduits.
Patrick Morrissey, Paul Nolan, Ted McCormack, Paul Johnston, Owen Naughton, Saheba Bhatnagar, and Laurence Gill
Hydrol. Earth Syst. Sci., 25, 1923–1941,Short summary
Lowland karst aquifers provide important wetland habitat resulting from seasonal flooding on the land surface. This flooding is controlled by surcharging of the karst system, which is very sensitive to changes in rainfall. This study investigates the predicted impacts of climate change on a lowland karst catchment in Ireland and highlights the relative vulnerability to future changing climate conditions of karst systems and any associated wetland habitats.
Liwen Wu, Jesus D. Gomez-Velez, Stefan Krause, Anders Wörman, Tanu Singh, Gunnar Nützmann, and Jörg Lewandowski
Hydrol. Earth Syst. Sci., 25, 1905–1921,Short summary
With a physically based model that couples flow and heat transport in hyporheic zones, the present study provides the first insights into the dynamics of hyporheic responses to the impacts of daily groundwater withdrawal and river temperature fluctuations, allowing for a better understanding of transient hyporheic exchange processes and hence an improved pumping operational scheme.
Andreas Wunsch, Tanja Liesch, and Stefan Broda
Hydrol. Earth Syst. Sci., 25, 1671–1687,
Jost Hellwig, Michael Stoelzle, and Kerstin Stahl
Hydrol. Earth Syst. Sci., 25, 1053–1068,Short summary
Potential future groundwater and baseflow drought hazards depend on systems' sensitivity to altered recharge conditions. With three generic scenarios, we found different sensitivities across Germany driven by hydrogeology. While changes in drought hazard due to seasonal recharge shifts will be rather low, a lengthening of dry spells could cause stronger responses in regions with slow groundwater response to precipitation, urging local water management to prepare for more severe droughts.
Martin J. Wells, Troy E. Gilmore, Natalie Nelson, Aaron Mittelstet, and John K. Böhlke
Hydrol. Earth Syst. Sci., 25, 811–829,Short summary
Groundwater in many agricultural areas contains high levels of nitrate, which is a concern for drinking water supplies. The rate at which nitrate moves through the subsurface is a critical piece of information for predicting how quickly groundwater nitrate levels may improve after agricultural producers change their approach to managing crop water and fertilizers. In this study, we explored a new statistical modeling approach to determine rates at which nitrate moves into and through an aquifer.
Karina Y. Gutierrez-Jurado, Daniel Partington, and Margaret Shanafield
Hydrol. Earth Syst. Sci. Discuss.,
Revised manuscript accepted for HESSShort summary
Understanding the hydrologic cycle in semi-arid landscapes includes knowing the physical processes that govern where and why rivers flow and dry within a given catchment. To gain this understanding, we put together a conceptual model of what processes we think are important, and then tested that model with numerical analysis. The results broadly confirmed our hypothesis that there are three distinct regions in our study catchment that contribute to streamflow generation in quite different ways.
Arnaud Duranel, Julian R. Thompson, Helene Burningham, Philippe Durepaire, Stéphane Garambois, Robert Wyns, and Hervé Cubizolle
Hydrol. Earth Syst. Sci., 25, 291–319,Short summary
Peat-forming wetlands (mires) provide multiple ecosystem services, which depend on peat remaining waterlogged. Using hydrological modelling, we show that, contrary to a common assumption, groundwater inflow can be a quantitatively important and functionally critical element of the water balance of mires in hard-rock upland and mountain areas. This influence is such that patterns of groundwater upwelling and seepage explain the spatial distribution of mires in the landscape.
Ming Wu, Jianfeng Wu, Jichun Wu, and Bill X. Hu
Hydrol. Earth Syst. Sci., 24, 5903–5917,Short summary
A new criterion (χi) is proposed to estimate representative elementary volume (REV) of a translucent material based on light transmission techniques. This study is essential for quantitative investigation of the scale effect of porous media and contaminant transformation. The fluid and contaminant migration and transform in porous media can be simulated accurately according to the REV estimation results using the light transmission technique and the appropriate criterion χi.
Marie-Amélie Pétré, Bernard Ladouche, Jean-Luc Seidel, Romain Hemelsdaël, Véronique de Montety, Christelle Batiot-Guilhe, and Claudine Lamotte
Hydrol. Earth Syst. Sci., 24, 5655–5672,Short summary
We studied the impact of occasional saltwater intrusions into the karst aquifer of the Balaruc peninsula (France). Using hydrogeological and geochemical data, this study shows that the hydraulic impact on the aquifer is rapid and of regional extent, whereas the geochemical impact is observed at the local scale and is temporally persistent. This research supports groundwater management by providing a better understanding of the hydrodynamics and recovery of the aquifer after saltwater intrusions.
Emmanuel Dubois, Joanna Doummar, Séverin Pistre, and Marie Larocque
Hydrol. Earth Syst. Sci., 24, 4275–4290,Short summary
The simulation of flow in a karst aquifer in a Mediterranean region using a semi-distributed linear reservoir model (geometry and parameterization) is calibrated and validated based on the analysis of high-resolution time series. The model is used to predict the effect of climatic variation. Although the spring is highly sensitive to rainfall variations, it is also resilient to warming temperature. Finally, this integrated conceptual method is reproducible for karst in semiarid regions.
Stephen R. Maples, Laura Foglia, Graham E. Fogg, and Reed M. Maxwell
Hydrol. Earth Syst. Sci., 24, 2437–2456,Short summary
In this study, we use a combination of local- and global-sensitivity analyses to evaluate the relative importance of (1) the configuration of subsurface alluvial geology and (2) the hydraulic properties of geologic facies on recharge processes. Results show that there is a large variation of recharge rates possible in a typical alluvial aquifer system and that the configuration proportion of sand and gravel deposits in the subsurface have a large impact on recharge rates.
Jian Song, Yun Yang, Xiaomin Sun, Jin Lin, Ming Wu, Jianfeng Wu, and Jichun Wu
Hydrol. Earth Syst. Sci., 24, 2323–2341,Short summary
We proposed a novel many-objective simulation-optimization framework for conjunctive use of surface water and groundwater in Yanqi Basin, northwest China. The management model involving socioeconomic and environmental objectives was constructed to explore optimal water-use schemes. Three runoff scenarios were then specified to quantify the effect of runoff reduction related to climate change on water management. Results provide Pareto-optimal solutions for basin-scale water management.
Miao Jing, Rohini Kumar, Falk Heße, Stephan Thober, Oldrich Rakovec, Luis Samaniego, and Sabine Attinger
Hydrol. Earth Syst. Sci., 24, 1511–1526,Short summary
This study investigates the response of regional groundwater system to the climate change under three global warming levels (1.5, 2, and 3 °C) in a central German basin. A comprehensive uncertainty analysis is also presented. This study indicates that the variability of responses increases with the amount of global warming, which might affect the cost of managing the groundwater system.
Harald Hofmann, Dean Newborn, Ian Cartwright, Dioni I. Cendón, and Matthias Raiber
Hydrol. Earth Syst. Sci., 24, 1293–1318,Short summary
Fresh groundwater (GW) on barrier islands is affected by GW use and precipitation variability. Mean residence times (MRTs) of GW on a sand barrier island were determined. They ranged from 37 years to more than 150 years for tritium and had a much larger range (modern to 5000 years) for carbon-14. Perched aquifer systems in the unsaturated zone and peat formations around wetlands are the most likely cause of longer MRTs, as they have a significant impact on regional recharge and flow diversion.
Christopher Vincent Henri, Thomas Harter, and Efstathios Diamantopoulos
Hydrol. Earth Syst. Sci., 24, 1189–1209,Short summary
Non-point source contaminations of aquifers are complex to model, predict and manage. This study uses numerical and stochastic methods to address the importance of key sources of spatial variability. We show that heterogeneity in recharge and contaminant loading does not significantly impact management metrics and could be simplified. Also, homogenizing physical properties has more impact on predictions, but can provide useful information on concentration statistics in a regional analysis.
Behzad Ataie-Ashtiani and Craig T. Simmons
Hydrol. Earth Syst. Sci., 24, 761–769,Short summary
We revisit and shed light on the textbook The Extraction of Hidden Waters by the Persian mathematician and engineer Abubakr Mohammad Karaji. Ground-breaking ideas and descriptions of hydrological and hydrogeological perceptions such as components of the hydrological cycle, groundwater quality and driving factors for groundwater flow were presented in the book. We speculate that Karaji's book is the first of its kind to provide a construction and maintenance manual for an engineering project.
Zhe Zhang, Yanping Li, Michael Barlage, Fei Chen, Gonzalo Miguez-Macho, Andrew Ireson, and Zhenhua Li
Hydrol. Earth Syst. Sci., 24, 655–672,Short summary
The groundwater regime in cold regions is strongly impacted by the soil freeze–thaw processes and semiarid climatic conditions. In this paper, we incorporate groundwater dynamics in the Noah-MP land surface model to simulate the water exchange between the unsaturated soil zone and an unconfined aquifer in the Prairie Pothole Region. The water table dynamics are reasonably simulated. The water budget of groundwater aquifer under current and future climate are also investigated.
Cornelia Wilske, Axel Suckow, Ulf Mallast, Christiane Meier, Silke Merchel, Broder Merkel, Stefan Pavetich, Tino Rödiger, Georg Rugel, Agnes Sachse, Stephan M. Weise, and Christian Siebert
Hydrol. Earth Syst. Sci., 24, 249–267,Short summary
Despite feeding several million people, the flow system and inter-aquifer communication within one of the major aquifer systems in Israel and the West Bank is still poorly understood. Applying a combination of inorganic elements, anthropogenic organic trace substances and classical environmental age-dating tracers like 3H, CFCs, SF6 and 36Cl / Cl, groundwater flow patterns, mixing end-members, transport times and recharge estimates have been obtained despite very limited data.
Joeri van Engelen, Jarno Verkaik, Jude King, Eman R. Nofal, Marc F. P. Bierkens, and Gualbert H. P. Oude Essink
Hydrol. Earth Syst. Sci., 23, 5175–5198,Short summary
The Nile Delta is an important agricultural area with a fast-growing population, relying increasingly on groundwater. However, saline groundwater extends far land-inward, rendering groundwater close to the coastal zone useless for consumption or agriculture. It normally is assumed that this is caused by mixing due to velocity differences, but here we show that it might also be caused by the coastline being located more land-inward 8000 years ago.
Julian Koch, Helen Berger, Hans Jørgen Henriksen, and Torben Obel Sonnenborg
Hydrol. Earth Syst. Sci., 23, 4603–4619,Short summary
This study explores novel modelling avenues using machine learning in combination with process-based models to predict the shallow water table at high spatial resolution. Due to climate change and anthropogenic impacts, the shallow groundwater is rising in many parts of the world. In order to adapt to risks induced by groundwater flooding, new modelling tools need to emerge. In this study, we found that machine learning is capable of reaching the required accuracy and resolution.
Donald W. Vasco, Joseph Doetsch, and Ralf Brauchler
Hydrol. Earth Syst. Sci., 23, 4541–4560,Short summary
This paper presents the application of a new approach for calculating the path of a pressure transient in a heterogeneous porous medium containing a slightly compressible fluid. Unlike previous asymptotic approaches, the expressions for the path and travel time are valid in the presence of rapid variations in material properties. The technique is applied to both synthetic transient pressure variations from a test example and actual field data from a field experiment in Widen, Switzerland.
Daniel Erdal and Olaf A. Cirpka
Hydrol. Earth Syst. Sci., 23, 3787–3805,Short summary
Assessing how sensitive uncertain model parameters are to observed data can be done by analyzing an ensemble of model simulations in which the parameters are varied. In subsurface modeling, this involves running heavy models. To reduce time wasted simulating models which show poor behavior, we use a fast polynomial model based on a simple parameter decomposition to approximate the behavior prior to full-model simulation. This largely reduces the cost for the global sensitivity analysis.
Wei Mao, Yan Zhu, Heng Dai, Ming Ye, Jinzhong Yang, and Jingwei Wu
Hydrol. Earth Syst. Sci., 23, 3481–3502,Short summary
A new quasi-3-D model was developed by coupling a soil water balance model with MODFLOW iteratively for regional-scale water flow modeling. The model was tested to be effective and efficient with well-maintained mass balance. A modeling framework was developed to organize the coupling scheme and to handle the pre- and post-processing information. The model is then used to evaluate groundwater recharge in a real-world application, which shows the model practicability in regional-scale problems.
Ning Li, Wolfgang Kinzelbach, Haitao Li, Wenpeng Li, and Fei Chen
Hydrol. Earth Syst. Sci., 23, 2823–2840,Short summary
Groundwater heads within an administrative unit are influenced not only by inside drivers, but also by outside drivers. To assess the efficiency of groundwater management of the administrative unit, we propose the decomposition of groundwater heads within the unit into inside and outside contributions by using three numerical groundwater models. The method is successfully demonstrated using Guantao County, China.
Zhenjiao Jiang, Dirk Mallants, Luk Peeters, Lei Gao, Camilla Soerensen, and Gregoire Mariethoz
Hydrol. Earth Syst. Sci., 23, 2561–2580,Short summary
Paleovalleys often form productive aquifers in the semiarid and arid areas. A methodology based on deep learning is introduced to automatically generate high-resolution 3-D paleovalley maps from low-resolution electrical conductivity data derived from airborne geophysical surveys. It is validated by borehole logs and the surface valley indices that the proposed method in this study provides an effective tool for regional-scale paleovalley mapping and groundwater exploration.
Nicholas D. Woodman, William G. Burgess, Kazi Matin Ahmed, and Anwar Zahid
Hydrol. Earth Syst. Sci., 23, 2461–2479,Short summary
We show that a conventional hydraulic understanding of groundwater level fluctuation is too simplistic for the extensive floodplains of Bangladesh and West Bengal. This is crucial because 150 million people of the region rely on groundwater for drinking and irrigation. We describe a more complex situation: the coupled hydro-mechanical action of surface water coming and going as the seasons change. Our model results will assist sustainable management of groundwater resources across the region.
Quanrong Wang and Hongbin Zhan
Hydrol. Earth Syst. Sci., 23, 2207–2223,Short summary
New multi-species reactive models of the four-phase SWPP test were presented considering the wellbore storages for both groundwater flow and solute transport and a finite-aquifer hydraulic diffusivity, which were ignored in previous studies. The models of the wellbore storage for solute transport were proposed based on the mass balance, and the sensitivity analysis and uniqueness analysis were employed to investigate the assumptions used in previous studies on the parameter estimation.
Wolfgang Dreybrodt and Franci Gabrovšek
Hydrol. Earth Syst. Sci., 23, 1995–2014,Short summary
Numerical models of wormhole formation in fractured porous media provide basic insights on the evolution of complex conduit systems in karst aquifers. In this work we use a time-propagating reactive flow model to explore the evolution of wormholes in a 2-D grid of fractures. We demonstrate physically meaningful mechanisms leading to the formation of individual wormholes and the competition between a set of evolving wormholes.
Ching-Sheng Huang, Ya-Hsin Tsai, Hund-Der Yeh, and Tao Yang
Hydrol. Earth Syst. Sci., 23, 1323–1337,Short summary
The study develops a new model describing head fluctuation induced by oscillatory pumping test (OPT) in an unconfined aquifer with effects of delayed gravity drainage (DGD) and initial condition regarding the hydraulic head prior to OPT. The DGD reduces to instantaneous gravity drainage when a dimensionless parameter exceeds 500. A pseudo-steady-state model excluding initial condition causes a time-shift from the actual transient model in predicting simple harmonic motion of head fluctuation.
Nimrod Inbar, Eliahu Rosenthal, Fabien Magri, Marwan Alraggad, Peter Möller, Akiva Flexer, Joseph Guttman, and Christian Siebert
Hydrol. Earth Syst. Sci., 23, 763–771,Short summary
In areas of enigmatic hydrology, water scarcity, and transboundary water resources, management strategies should rely on comprehensive modeling which must be based on realistic geometry, including all relevant structural features. Based on available geophysical and geological data, a new faulting pattern in the Lower Yarmouk Gorge is suggested as a basis for hydrogeological modeling. Furthermore, unexpected pull-apart basin rim fault evolution is discussed in the context of tectonic collision.
I-Hsien Lee, Chuen-Fa Ni, Fang-Pang Lin, Chi-Ping Lin, and Chien-Chung Ke
Hydrol. Earth Syst. Sci., 23, 19–34,Short summary
Few studies focused on the direct solution of the ADE for 3-D DFNs. The study is the first to solve the ADE and focus on assessing the velocity uncertainty in 3-D DFNs. The velocity uncertainty shows a limited range of influence close to the mean diameter of a fracture. The information is useful for engineering designs at sites with fractured rocks. We quantified that the tracer test in wells might lead to the overestimation of mean concentration and induce high uncertainty in fractured media.
Qiyu Chen, Gregoire Mariethoz, Gang Liu, Alessandro Comunian, and Xiaogang Ma
Hydrol. Earth Syst. Sci., 22, 6547–6566,Short summary
One of the critical issues in MPS simulation is the difficulty in obtaining a credible 3-D training image. We propose an MPS-based 3-D reconstruction method on the basis of 2-D cross sections, making 3-D training images unnecessary. The main advantages of this approach are the high computational efficiency and a relaxation of the stationarity assumption. The results, in comparison with previous MPS methods, show better performance in portraying anisotropy characteristics and in CPU cost.
Rogier Westerhoff, Paul White, and Gonzalo Miguez-Macho
Hydrol. Earth Syst. Sci., 22, 6449–6472,Short summary
Our study improved a global-scale groundwater model to build the first nationwide estimate of the water table surface in New Zealand. By identifying the main alluvial aquifers with high spatial detail, we showed that this model can help better delineate aquifer boundaries. In catchment studies we demonstrated excellent correlation with ground observations and provided water table estimates where data were sparse and across regions, which could help solve trans-boundary issues between catchments.
Yonatan Ganot, Ran Holtzman, Noam Weisbrod, Anat Bernstein, Hagar Siebner, Yoram Katz, and Daniel Kurtzman
Hydrol. Earth Syst. Sci., 22, 6323–6333,Short summary
In recent years, surpluses of desalinated seawater (DSW) are stored in the Israeli coastal aquifer. We monitor DSW spread in the aquifer using the difference between isotope composition of reverse-osmosis DSW and natural fresh water, which simplifies the system to two distinct end-members. A hydrogeological flow and transport model is used to demonstrate the robustness of this simplification, predict the future spread of DSW in the aquifer and mixing in wells, and estimate DSW recovery efficacy.
Claudia Cherubini, Nicola Pastore, Dimitra Rapti, and Concetta I. Giasi
Hydrol. Earth Syst. Sci., 22, 5211–5225,Short summary
The present study investigates the hotspots of groundwater contamination in the industrial area of Modugno (Bari – southern Italy) where the limestone aquifer has a fractured and karstic nature. The characterization of the dynamics of fluid flow and solute transport in fractured aquifers is of broad concern worldwide, especially when considered in relation to water resource exploitation, because fractured aquifers serve as the primary source of drinking water for many areas of the world.
Rena Meyer, Peter Engesgaard, Klaus Hinsby, Jan A. Piotrowski, and Torben O. Sonnenborg
Hydrol. Earth Syst. Sci., 22, 4843–4865,
Khabat Khosravi, Mahdi Panahi, and Dieu Tien Bui
Hydrol. Earth Syst. Sci., 22, 4771–4792,
Yong Xiao, Jingli Shao, Shaun K. Frape, Yali Cui, Xueya Dang, Shengbin Wang, and Yonghong Ji
Hydrol. Earth Syst. Sci., 22, 4381–4400,
Martin Suchy, Leonard I. Wassenaar, Gwyn Graham, and Bernie Zebarth
Hydrol. Earth Syst. Sci., 22, 4267–4279,Short summary
Groundwater contamination by nitrate is a global problem. Understanding sources and transformations of nitrate in aquifers is key to inform remediation, and isotopes of nitrate (15N, 18O) are used for this. Most isotopic investigations are synoptic studies. Here we show that high-frequency nitrate isotope sampling of aquifer recharge better reveals the impact of short-term changes in land use practice or climatic conditions that need to be taken into account when considering long-term trends.
Elliott P. Morrill and Joseph F. Becker
Hydrol. Earth Syst. Sci., 22, 3761–3775,Short summary
The goal of the paper was to develop a method to identify the length and severity of flood events for (re)insurance and risk management usage. We took publically available nationwide discharge data from the USGS to apply our method to. We were able to define a peak and a threshold for each individual site, which served as the basis of our method. The hope was that the method would help replace the current standard hours clause and hopefully improve how flood insurance is covered internationally.
David Pulido-Velazquez, Arianna Renau-Pruñonosa, Carlos Llopis-Albert, Ignacio Morell, Antonio-Juan Collados-Lara, Javier Senent-Aparicio, and Leticia Baena-Ruiz
Hydrol. Earth Syst. Sci., 22, 3053–3074,Short summary
We have developed a method to assess hydrological impacts of future potential global change (GC) scenarios in a coastal aquifer, the Plana Oropesa-Torreblanca aquifer. Quantity and quality issues are simultaneously considered. It includes the generation of potential GC scenarios involving different sources of uncertainty and an integrated modelling framework to assess their impacts. It is a valuable tool to improve our aquifer knowledge and to help in the analysis of management strategies.
Martina Siena and Monica Riva
Hydrol. Earth Syst. Sci., 22, 2971–2985,Short summary
The development of sustainable strategies for groundwater resources exploitation in coastal regions is subordinated to the characterization of seawater intrusion (SWI) phenomena. We develop a numerical model tailored to a real coastal aquifer to investigate quantitatively the joint effects of hydraulic properties heterogeneity and pumping configuration on saltwater inland penetration and saltwater–freshwater mixing. Our results allowed identifying efficient scenarios for the reduction of SWI.
Katherine M. Ransom, Andrew M. Bell, Quinn E. Barber, George Kourakos, and Thomas Harter
Hydrol. Earth Syst. Sci., 22, 2739–2758,Short summary
We estimated a probability distribution of nitrogen loading rates for crop and land-use groups from regional groundwater data. Water & natural land use had the lowest estimated rates, while dairy land use had the highest. Most results compare favorably to previous estimates, though mass balance estimates for several crop groups were higher than our model estimates. The information can provide a better assessment of land-use impacts to water quality absent information on farm nutrient management.
Sebastian Huizer, Max Radermacher, Sierd de Vries, Gualbert H. P. Oude Essink, and Marc F. P. Bierkens
Hydrol. Earth Syst. Sci., 22, 1065–1080,Short summary
For a large beach nourishment called the Sand Engine we have examined the impact of groundwater recharge, tides, storm surges, and geomorphological changes on the growth of the fresh groundwater resources between 2011 and 2016. With detailed model simulations of these coastal processes we were able to get a good match with field measurements, and demonstrated the importance of wave runup and coastal erosion in studies on fresh groundwater in such dynamic coastal environments.
Pietro Teatini, Giovanni Isotton, Stefano Nardean, Massimiliano Ferronato, Annamaria Mazzia, Cristina Da Lio, Luca Zaggia, Debora Bellafiore, Massimo Zecchin, Luca Baradello, Francisco Cellone, Fabiana Corami, Andrea Gambaro, Giovanni Libralato, Elisa Morabito, Annamaria Volpi Ghirardini, Riccardo Broglia, Stefano Zaghi, and Luigi Tosi
Hydrol. Earth Syst. Sci., 21, 5627–5646,Short summary
We investigate the effects of digging a navigable canal on the hydrogeological system underlying a coastal lagoon. The research has been promoted by the Venice Water Authority, which is investigating different possibilities to avoid the passage of large cruise ships through the historic center of Venice, Italy. Numerical simulations supported by a proper hydrogeological characterization show that the exchange of water and contaminants from the subsurface and surface systems will be significant.
Abadi, M., Agarwal, A., Barham, P., Brevdo, E., Chen, Z., Citro, C., Corrado, G. S., Davis, A., Dean, J., Devin, M., Ghemawat, S., Goodfellow, I. J., Harp, A., Irving, G., Isard, M., Jia, Y., Józefowicz, R., Kaiser, L., Kudlur, M., Levenberg, J., Mané, D., Monga, R., Moore, S., Murray, D. G., Olah, C., Schuster, M., Shlens, J., Steiner, B., Sutskever, I., Talwar, K., Tucker, P. A., Vanhoucke, V., Vasudevan, V., Viégas, F. B., Vinyals, O., Warden, P., Wattenberg, M., Wicke, M., Yu, Y., and Zheng, X.: TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems, CoRR, abs/1603.04467, http://arxiv.org/abs/1603.04467, 2016. a
Adamowski, J., Adamowski, K., and Prokoph, A.: A spectral analysis based methodology to detect climatological influences on daily urban water demand, Math. Geosci., 45, 49–68, 2013. a
Adhikari, S., Belasco, E. J., and Knight, T. O.: Spatial producer heterogeneity in crop insurance product decisions within major corn producing states, Agric. Financ. Rev., 70, 66–78, 2010. a
Ahire, J.: Artificial Neural Networks: the Brain behind AI, Lulu. com, ISBN 1980483671, 9781980483670, 2018. a
Amaranto, A., Munoz-Arriola, F., Solomatine, D., and Corzo, G.: A spatially enhanced data-driven multimodel to improve semiseasonal groundwater forecasts in the High Plains aquifer, USA, Water Resour. Res., 55, 5941–5961, 2019. a
Barnes, S. L.: A technique for maximizing details in numerical weather map analysis, J. Appl. Meteorol., 3, 396–409, 1964. a
Bennasar, M., Hicks, Y., and Setchi, R.: Feature selection using joint mutual information maximisation, Expert Syst. Appl., 42, 8520–8532, 2015. a
Bi, J.: A review of statistical methods for determination of relative importance of correlated predictors and identification of drivers of consumer liking, J. Sens. Stud., 27, 87–101, 2012. a
Breiman, L.: Random forests, Mach. Learn., 45, 5–32, 2001. a
Chollet, F. et al.: Keras, available at: https://keras.io (last access: 20 January 2020), 2015. a
Daliakopoulos, I. N., Coulibaly, P., and Tsanis, I. K.: Groundwater level forecasting using artificial neural networks, J. Hydrol., 309, 229–240, 2005. a
Farthing, M. W., Kees, C. E., and Miller, C. T.: Mixed finite element methods and higher order temporal approximations for variably saturated groundwater flow, Adv. Water Resour., 26, 373–394, 2003. a
Gleeson, T., Alley, W. M., Allen, D. M., Sophocleous, M. A., Zhou, Y., Taniguchi, M., and VanderSteen, J.: Towards sustainable groundwater use: setting long-term goals, backcasting, and managing adaptively, Groundwater, 50, 19–26, 2012. a
Gómez-Hernández, J. J. and Gorelick, S. M.: Effective groundwater model parameter values: Influence of spatial variability of hydraulic conductivity, leakance, and recharge, Water Resour. Res., 25, 405–419, 1989. a
Grömping, U.: Variable importance assessment in regression: linear regression versus random forest, Am. Stat., 63, 308–319, 2009. a
Gutzler, D. S. and Robbins, T. O.: Climate variability and projected change in the western United States: regional downscaling and drought statistics, Clim. Dyn., 37, 835–849, 2011. a
Isaaks, E. H. and Srivastava, M. R.: Applied geostatistics, 551.72 ISA, Oxford University Press, ISBN 0195050134, 9780195050134, 1989. a
Kleijnen, J. P. C. and Helton, J. C.: Statistical analyses of scatterplots to identify important factors in large-scale simulations, 1: Review and comparison of techniques, Reliab. Eng. Syst. Saf., 65, 147–185, 1999. a
Kursa, M. B., Jankowski, A., and Rudnicki, W. R.: Boruta – a system for feature selection, Fundam. Informaticae, 101, 271–285, 2010. a
LeCun, Y. A., Bottou, L., Orr, G. B., and Müller, K.-R.: Efficient backprop, in: Neural networks: Tricks of the trade, pp. 9–48, Springer, 2012. a
Liu, D., Li, G., Fu, Q., Li, M., Liu, C., Faiz, M. A., Khan, M. I., Li, T., and Cui, S.: Application of Particle Swarm Optimization and Extreme Learning Machine Forecasting Models for Regional Groundwater Depth Using Nonlinear Prediction Models as Preprocessor, J. Hydrol. Eng., 23, 4018052, 2018. a
Maier, H. R. and Dandy, G. C.: Neural network based modelling of environmental variables: a systematic approach, Math. Comput. Model., 33, 669–682, 2001. a
Marston, L., Konar, M., Cai, X., and Troy, T. J.: Virtual groundwater transfers from overexploited aquifers in the United States, P. Natl. Acad. Sci., 112, 8561–8566, 2015. a
Meyer, P. E.: infotheo: Information-Theoretic Measures, available at: https://cran.r-project.org/package=infotheo (last access: 2 February 2020), 2014. a
Mickens, R. E.: Advances in the Applications of Nonstandard Finite Diffference Schemes, World Scientific, 2005. a
Mohanty, S., Jha, M. K., Raul, S., Panda, R., and Sudheer, K.: Using artificial neural network approach for simultaneous forecasting of weekly groundwater levels at multiple sites, Water Resour. Manag., 29, 5521–5532, 2015. a
Nizar Shamsuddin, M. K., Mohd Kusin, F., Sulaiman, W. N. A., Ramli, M. F., Tajul Baharuddin, M. F., and Adnan, M. S.: Forecasting of Groundwater Level using Artificial Neural Network by incorporating river recharge and river bank infiltration, in: MATEC Web Conf., 103, 04007, https://doi.org/10.1051/matecconf/201710304007, 2017. a
Oreskes, N., Shrader-Frechette, K., and Belitz, K.: Verification, validation, and confirmation of numerical models in the earth sciences, Science, 263, 641–646, 1994. a
Roshni, T., Jha, M. K., Deo, R. C., and Vandana, A.: Development and Evaluation of Hybrid Artificial Neural Network Architectures for Modeling Spatio-Temporal Groundwater Fluctuations in a Complex Aquifer System, Water Resour. Manag., 33, 2381–2397, https://doi.org/10.1007/s11269-019-02253-4, 2019. a
Siroky, D. S.: Navigating random forests and related advances in algorithmic modeling, Statistics Surveys, 3, 147–163, 2009. a
Thornthwaite, C. W. and Mather, J. R.: Instructions and tables for computing potential evapotranspiration and the water balance, Tech. rep., Centerton, 1957. a
Thornton, P. K., Ericksen, P. J., Herrero, M., and Challinor, A. J.: Climate variability and vulnerability to climate change: a review, Glob. Chang. Biol., 20, 3313–3328, 2014. a
Trichakis, I. C., Nikolos, I. K., and Karatzas, G. P.: Artificial neural network (ANN) based modeling for karstic groundwater level simulation, Water Resour. Manag., 25, 1143–1152, 2011. a
TWDB: http://www.twdb.texas.gov/groundwater/data/ (last access: 20 October 2019), 2019. a
Uddameri, V. and Andruss, T.: A GIS-based multi-criteria decision-making approach for establishing a regional-scale groundwater monitoring, Environ. earth Sci., 71, 2617–2628, 2014. a
Uddameri, V. and Reible, D.: Food-energy-water nexus to mitigate sustainability challenges in a groundwater reliant agriculturally dominant environment (GRADE), Environ. Prog. Sustain. Energy, 37, 21–36, 2018. a
USDA-NASS: 2012 Census of Agriculture US, available at: https://www.nass.usda.gov/Publications/AgCensus/2012/Full_Report/Volume_1,_Chapter_1_US/usv1.pdf (last access: 25 January 2020), 2012. a
Varouchakis, E. A., Theodoridou, P. G., and Karatzas, G. P.: Spatiotemporal geostatistical modeling of groundwater levels under a Bayesian framework using means of physical background, J. Hydrol., 575, 487–498, 2019. a
Vergara, J. R. and Estévez, P. A.: A review of feature selection methods based on mutual information, Neural Comput. Appl., 24, 175–186, 2014. a
Vicente-Serrano, S. M., Beguería, S., and López-Moreno, J. I.: A multiscalar drought index sensitive to global warming: the standardized precipitation evapotranspiration index, J. Clim., 23, 1696–1718, 2010. a
Whittemore, D. O., Butler Jr., J. J., and Wilson, B. B.: Assessing the major drivers of water-level declines: New insights into the future of heavily stressed aquifers, Hydrol. Sci. J., 61, 134–145, 2016. a
Zambrano-Bigiarini, M.: hydroGOF: Goodness-of-fit functions for comparison of simulated and observed hydrological time series, R Packag. version 0.3-8, 2014. a
Zambreski, Z. T., Lin, X., Aiken, R. M., Kluitenberg, G. J., and Pielke Sr., R. A.: Identification of hydroclimate subregions for seasonal drought monitoring in the US Great Plains, J. Hydrol., 567, 370–381, 2018. a
While artificial neural networks (ANNs) have been used to forecast groundwater levels at single wells, they have not been constructed to forecast hydraulic heads in both space and time. This seminal study presents a modeling framework, guided by the governing physical laws, for building an integrated space–time ANN (IST–ANN) model for regional groundwater level predictions. IST–ANN shows promise for parsimoniously modeling regional-scale groundwater levels using available surrogate information.
While artificial neural networks (ANNs) have been used to forecast groundwater levels at single...