Knowledge of the groundwater recharge rate determines whether aquifer use is sustainable. However, accurately measuring recharge globally presents significant challenges due to the complexity of subsurface processes and the lack of direct observational methods. This study addresses these challenges by developing a methodology that integrates satellite data, numerical models, and machine learning to estimate groundwater recharge globally. The methodology involves two steps. First, we run a numerical model, Hydrus-1D, to simulate soil moisture fluxes in the unsaturated zone by solving the Richards equation in the vertical direction for 235 different points representing various climates and soil types across the globe. Second, using Hydrus-1D inputs and outputs, we train a supervised ensemble machine-learning model, specifically a Gaussian Process Regression model, as an emulator to mimic Hydrus-1D. This enables us to process satellite observations efficiently to estimate annual recharge flux globally. Inputs for the model include NASA's SMAP soil moisture and GPM precipitation observations, ERA5 climate reanalysis data, and soil hydraulic properties. Rainfall, unsaturated hydraulic conductivity, and soil moisture are identified as the most significant predictors of groundwater recharge. The approach effectively captures global recharge patterns, particularly in regions with high rainfall, though it shows some limitations in arid areas with minimal recharge and heavily irrigated areas. We confirm the reasonableness of recharge estimates by comparing them with observed changes in subsurface water storage from the GRACE satellite mission. The method effectively captures the observed trends in water storage, demonstrating the model's capability to estimate recharge using large-scale satellite and reanalysis data.
{"title":"Integrating Satellite Retrievals, Numerical Models, and Machine Learning for Global Groundwater Recharge Estimation","authors":"M. E. Soylu, D. Entekhabi, R. L. Bras","doi":"10.1029/2025wr040312","DOIUrl":"https://doi.org/10.1029/2025wr040312","url":null,"abstract":"Knowledge of the groundwater recharge rate determines whether aquifer use is sustainable. However, accurately measuring recharge globally presents significant challenges due to the complexity of subsurface processes and the lack of direct observational methods. This study addresses these challenges by developing a methodology that integrates satellite data, numerical models, and machine learning to estimate groundwater recharge globally. The methodology involves two steps. First, we run a numerical model, Hydrus-1D, to simulate soil moisture fluxes in the unsaturated zone by solving the Richards equation in the vertical direction for 235 different points representing various climates and soil types across the globe. Second, using Hydrus-1D inputs and outputs, we train a supervised ensemble machine-learning model, specifically a Gaussian Process Regression model, as an emulator to mimic Hydrus-1D. This enables us to process satellite observations efficiently to estimate annual recharge flux globally. Inputs for the model include NASA's SMAP soil moisture and GPM precipitation observations, ERA5 climate reanalysis data, and soil hydraulic properties. Rainfall, unsaturated hydraulic conductivity, and soil moisture are identified as the most significant predictors of groundwater recharge. The approach effectively captures global recharge patterns, particularly in regions with high rainfall, though it shows some limitations in arid areas with minimal recharge and heavily irrigated areas. We confirm the reasonableness of recharge estimates by comparing them with observed changes in subsurface water storage from the GRACE satellite mission. The method effectively captures the observed trends in water storage, demonstrating the model's capability to estimate recharge using large-scale satellite and reanalysis data.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"15 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2026-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147506478","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Wenli Zhao, Jianing Fang, Tao Yang, Xu Lian, Alexander J. Winkler, Fubao Sun, Pierre Gentine
Estimating daily snow water equivalent (SWE) is critical for hydrological and climate applications, yet physical models often struggle to represent SWE, especially its interannual anomalies. In this study, we developed a hybrid physics-guided machine learning (ML) model (hybrid model), by augmenting the Community Land Model 4.0 SWE simulations with a long short-term memory (LSTM) network. The model is trained using the GlobSnow v3.0 data set and forced with meteorological data to estimate daily SWE at 0.5° over the Northern Hemisphere (NH). Our results demonstrate that the hybrid model significantly outperforms both the standalone physical and pure ML models in predicting SWE magnitude, timing, and anomalies, especially in complex mountainous regions. Explainable ML analyses suggest that the hybrid approach leverages the snow-related physics while effectively utilizing observational data to enhance predictive accuracy. Moreover, we identify a widespread climate memory effect influencing SWE predictions across the NH, with memory-dominant extreme events leading to greater SWE losses or gains relative to the average impacts of all extreme events, including those without strong memory effects. These findings underscore the hybrid model's ability to correct memory-related biases that are not fully captured in current land surface models. Overall, our study highlights the value of hybrid modeling for improving SWE simulations and its potential as an alternative snow emulator within existing land surface models.
{"title":"Observation-Constrained Physical Snow Water Equivalent Simulations Using a Physics-Guided Machine Learning Approach","authors":"Wenli Zhao, Jianing Fang, Tao Yang, Xu Lian, Alexander J. Winkler, Fubao Sun, Pierre Gentine","doi":"10.1029/2025wr041406","DOIUrl":"https://doi.org/10.1029/2025wr041406","url":null,"abstract":"Estimating daily snow water equivalent (SWE) is critical for hydrological and climate applications, yet physical models often struggle to represent SWE, especially its interannual anomalies. In this study, we developed a hybrid physics-guided machine learning (ML) model (hybrid model), by augmenting the Community Land Model 4.0 SWE simulations with a long short-term memory (LSTM) network. The model is trained using the GlobSnow v3.0 data set and forced with meteorological data to estimate daily SWE at 0.5° over the Northern Hemisphere (NH). Our results demonstrate that the hybrid model significantly outperforms both the standalone physical and pure ML models in predicting SWE magnitude, timing, and anomalies, especially in complex mountainous regions. Explainable ML analyses suggest that the hybrid approach leverages the snow-related physics while effectively utilizing observational data to enhance predictive accuracy. Moreover, we identify a widespread climate memory effect influencing SWE predictions across the NH, with memory-dominant extreme events leading to greater SWE losses or gains relative to the average impacts of all extreme events, including those without strong memory effects. These findings underscore the hybrid model's ability to correct memory-related biases that are not fully captured in current land surface models. Overall, our study highlights the value of hybrid modeling for improving SWE simulations and its potential as an alternative snow emulator within existing land surface models.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"1 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2026-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147492826","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Hydrodynamic and morphodynamic forces interacting across the sediment-water interface control the biogeochemistry in the hyporheic zone. When investigating the redox zonation within streambeds, dissolved oxygen (O2) is considered a key solute to the understanding of river ecosystems. However, no field studies have measured the spatiotemporal O2 distribution linked to bedform celerity induced by changes in stream water velocity. Therefore, we developed and tested an innovative in situ setup in the River Erpe, Germany. The setup combines a planar O2 optode and O2 flow-through cells for eight sediment depths to capture the variability in O2 dynamics, and a laser scanner to capture bedform morphodynamics, which was used to calculate bedform celerity. The setup was tested under different stream water velocities between 0.1 and 0.5 m/s. We found that O2 patterns in the streambed depend on stream water velocity. At low velocities, bedforms were stationary and a stable redox zonation with limited O2 penetration in the streambed (up to 4 cm) was observed. As we increased the velocity up to 0.3 m/s, the spatiotemporal variability of O2 distribution across the bedform increased, with anoxic patches moving along the migrating bedforms. At the highest velocity tested (0.5 m/s), the sediment bed was constantly oxygenated with deeper O2 penetration as compared to slower velocities. The present study provides proof of concept for in situ O2 measurements in small rivers, which helps to refine laboratory and mesocosm experiments, improve the knowledge of the processes involved in natural environments, and develop more sustainable river management strategies.
{"title":"A Novel In Situ Experimental Setup for Studying the Impact of Bedform Celerity on 2D Oxygen Distribution in the Hyporheic Zone of Streams","authors":"Alejandra Villa, Hanna Schulz, Hauke Dämpfling, Stephanie Spahr, Shai Arnon, Jörg Lewandowski","doi":"10.1029/2025wr041376","DOIUrl":"https://doi.org/10.1029/2025wr041376","url":null,"abstract":"Hydrodynamic and morphodynamic forces interacting across the sediment-water interface control the biogeochemistry in the hyporheic zone. When investigating the redox zonation within streambeds, dissolved oxygen (O<sub>2</sub>) is considered a key solute to the understanding of river ecosystems. However, no field studies have measured the spatiotemporal O<sub>2</sub> distribution linked to bedform celerity induced by changes in stream water velocity. Therefore, we developed and tested an innovative in situ setup in the River Erpe, Germany. The setup combines a planar O<sub>2</sub> optode and O<sub>2</sub> flow-through cells for eight sediment depths to capture the variability in O<sub>2</sub> dynamics, and a laser scanner to capture bedform morphodynamics, which was used to calculate bedform celerity. The setup was tested under different stream water velocities between 0.1 and 0.5 m/s. We found that O<sub>2</sub> patterns in the streambed depend on stream water velocity. At low velocities, bedforms were stationary and a stable redox zonation with limited O<sub>2</sub> penetration in the streambed (up to 4 cm) was observed. As we increased the velocity up to 0.3 m/s, the spatiotemporal variability of O<sub>2</sub> distribution across the bedform increased, with anoxic patches moving along the migrating bedforms. At the highest velocity tested (0.5 m/s), the sediment bed was constantly oxygenated with deeper O<sub>2</sub> penetration as compared to slower velocities. The present study provides proof of concept for in situ O<sub>2</sub> measurements in small rivers, which helps to refine laboratory and mesocosm experiments, improve the knowledge of the processes involved in natural environments, and develop more sustainable river management strategies.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"20 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2026-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147492772","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Zhaohui Deng, Jiaxue Wu, Huan Liu, Jie Ren, Zhixin Cao
A deep understanding of flow regulation in river bifurcations is critical for downstream river management and flood control. The deltaic river network frequently contains the H-shaped network node, where two main channels are linked together via a connecting channel, combining both bifurcation and confluence. However, flow structures and their control on downstream discharge regulation of the H-shaped node remain unclear in river hydraulics. This study employed a high-resolution hydrodynamic model to simulate the flow dynamics of the H-shaped node in the Pearl River network, revealing three major findings: (a) the H-shaped node produces a more uniform discharge distribution between downstream channel branches than a single bifurcation; (b) the connecting channel plays a core role in downstream discharge regulation. When the upstream discharge ratio (UDR) exceeds a threshold (ranging between 3 and 4 in this study), the surface elevation difference between two junctions of the connecting channel causes the through flow to reverse its flow direction; (c) the switching of flow direction results in a dynamic transition between confluence-type and bifurcation-type flow structures at the two junctions. For <span data-altimg="/cms/asset/c318fab2-e2b4-4da7-8b04-d2804e657bdf/wrcr70802-math-0001.png"></span><mjx-container ctxtmenu_counter="133" ctxtmenu_oldtabindex="1" jax="CHTML" role="application" sre-explorer- style="font-size: 103%; position: relative;" tabindex="0"><mjx-math aria-hidden="true" location="graphic/wrcr70802-math-0001.png"><mjx-semantics><mjx-mrow data-semantic-children="0,4" data-semantic-collapsed="(6 (c 5) 0 4)" data-semantic- data-semantic-role="text" data-semantic-speech="UDR greater than 4" data-semantic-type="punctuated"><mjx-mtext data-semantic-annotation="clearspeak:unit" data-semantic-font="normal" data-semantic- data-semantic-parent="6" data-semantic-role="unknown" data-semantic-type="text"><mjx-c></mjx-c><mjx-c></mjx-c><mjx-c></mjx-c></mjx-mtext><mjx-mrow data-semantic-children="3,2" data-semantic-content="1" data-semantic- data-semantic-parent="6" data-semantic-role="inequality" data-semantic-type="relseq"><mjx-mrow data-semantic- data-semantic-parent="4" data-semantic-role="unknown" data-semantic-type="empty"></mjx-mrow><mjx-mo data-semantic- data-semantic-operator="relseq,>" data-semantic-parent="4" data-semantic-role="inequality" data-semantic-type="relation" rspace="5" space="5"><mjx-c></mjx-c></mjx-mo><mjx-mn data-semantic-annotation="clearspeak:simple" data-semantic-font="normal" data-semantic- data-semantic-parent="4" data-semantic-role="integer" data-semantic-type="number"><mjx-c></mjx-c></mjx-mn></mjx-mrow></mjx-mrow></mjx-semantics></mjx-math><mjx-assistive-mml display="inline" unselectable="on"><math altimg="urn:x-wiley:00431397:media:wrcr70802:wrcr70802-math-0001" display="inline" location="graphic/wrcr70802-math-0001.png" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow data-semantic-="" data-semanti
{"title":"Flow Structures and Their Controls on Downstream Discharge Regulation for a Combined Confluence and Bifurcation Network Node","authors":"Zhaohui Deng, Jiaxue Wu, Huan Liu, Jie Ren, Zhixin Cao","doi":"10.1029/2025wr041762","DOIUrl":"https://doi.org/10.1029/2025wr041762","url":null,"abstract":"A deep understanding of flow regulation in river bifurcations is critical for downstream river management and flood control. The deltaic river network frequently contains the H-shaped network node, where two main channels are linked together via a connecting channel, combining both bifurcation and confluence. However, flow structures and their control on downstream discharge regulation of the H-shaped node remain unclear in river hydraulics. This study employed a high-resolution hydrodynamic model to simulate the flow dynamics of the H-shaped node in the Pearl River network, revealing three major findings: (a) the H-shaped node produces a more uniform discharge distribution between downstream channel branches than a single bifurcation; (b) the connecting channel plays a core role in downstream discharge regulation. When the upstream discharge ratio (UDR) exceeds a threshold (ranging between 3 and 4 in this study), the surface elevation difference between two junctions of the connecting channel causes the through flow to reverse its flow direction; (c) the switching of flow direction results in a dynamic transition between confluence-type and bifurcation-type flow structures at the two junctions. For <span data-altimg=\"/cms/asset/c318fab2-e2b4-4da7-8b04-d2804e657bdf/wrcr70802-math-0001.png\"></span><mjx-container ctxtmenu_counter=\"133\" ctxtmenu_oldtabindex=\"1\" jax=\"CHTML\" role=\"application\" sre-explorer- style=\"font-size: 103%; position: relative;\" tabindex=\"0\"><mjx-math aria-hidden=\"true\" location=\"graphic/wrcr70802-math-0001.png\"><mjx-semantics><mjx-mrow data-semantic-children=\"0,4\" data-semantic-collapsed=\"(6 (c 5) 0 4)\" data-semantic- data-semantic-role=\"text\" data-semantic-speech=\"UDR greater than 4\" data-semantic-type=\"punctuated\"><mjx-mtext data-semantic-annotation=\"clearspeak:unit\" data-semantic-font=\"normal\" data-semantic- data-semantic-parent=\"6\" data-semantic-role=\"unknown\" data-semantic-type=\"text\"><mjx-c></mjx-c><mjx-c></mjx-c><mjx-c></mjx-c></mjx-mtext><mjx-mrow data-semantic-children=\"3,2\" data-semantic-content=\"1\" data-semantic- data-semantic-parent=\"6\" data-semantic-role=\"inequality\" data-semantic-type=\"relseq\"><mjx-mrow data-semantic- data-semantic-parent=\"4\" data-semantic-role=\"unknown\" data-semantic-type=\"empty\"></mjx-mrow><mjx-mo data-semantic- data-semantic-operator=\"relseq,>\" data-semantic-parent=\"4\" data-semantic-role=\"inequality\" data-semantic-type=\"relation\" rspace=\"5\" space=\"5\"><mjx-c></mjx-c></mjx-mo><mjx-mn data-semantic-annotation=\"clearspeak:simple\" data-semantic-font=\"normal\" data-semantic- data-semantic-parent=\"4\" data-semantic-role=\"integer\" data-semantic-type=\"number\"><mjx-c></mjx-c></mjx-mn></mjx-mrow></mjx-mrow></mjx-semantics></mjx-math><mjx-assistive-mml display=\"inline\" unselectable=\"on\"><math altimg=\"urn:x-wiley:00431397:media:wrcr70802:wrcr70802-math-0001\" display=\"inline\" location=\"graphic/wrcr70802-math-0001.png\" xmlns=\"http://www.w3.org/1998/Math/MathML\"><semantics><mrow data-semantic-=\"\" data-semanti","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"11 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2026-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147465942","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Empirical baseflow filters are widely used for baseflow separation. These filters rely on ad-hoc parameters that introduce significant uncertainties in the calculation. A recent study by Mei et al. (2024, https://doi.org/10.1029/2023wr036386) optimized these parameters using environmental tracer data for 1,100 catchments across the Contiguous United States (CONUS). However, optimized parameters are unavailable for most CONUS catchments lacking tracer data. To address this gap, we developed regionalization models for the filter parameters using the random forest (RF) algorithm and 82 catchment-scale characteristics, including geomorphology, climate, soil properties, and human activities. We demonstrated this approach for the block length parameter <span data-altimg="/cms/asset/f43cb71a-820f-4b8b-a8c0-c09af9c296ff/wrcr70795-math-0001.png"></span><mjx-container ctxtmenu_counter="436" ctxtmenu_oldtabindex="1" jax="CHTML" role="application" sre-explorer- style="font-size: 103%; position: relative;" tabindex="0"><mjx-math aria-hidden="true" location="graphic/wrcr70795-math-0001.png"><mjx-semantics><mjx-mrow><mjx-mi data-semantic-annotation="clearspeak:simple" data-semantic-font="italic" data-semantic- data-semantic-role="latinletter" data-semantic-speech="upper N" data-semantic-type="identifier"><mjx-c></mjx-c></mjx-mi></mjx-mrow></mjx-semantics></mjx-math><mjx-assistive-mml display="inline" unselectable="on"><math altimg="urn:x-wiley:00431397:media:wrcr70795:wrcr70795-math-0001" display="inline" location="graphic/wrcr70795-math-0001.png" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi data-semantic-="" data-semantic-annotation="clearspeak:simple" data-semantic-font="italic" data-semantic-role="latinletter" data-semantic-speech="upper N" data-semantic-type="identifier">N</mi></mrow>$N$</annotation></semantics></math></mjx-assistive-mml></mjx-container> of the smooth minima baseflow filter, one of the optimized filters in Mei et al.’s study, across 855 catchments. Our results show that the prediction of <span data-altimg="/cms/asset/816eb4e8-afca-4152-a4ac-77f24d01f79e/wrcr70795-math-0002.png"></span><mjx-container ctxtmenu_counter="437" ctxtmenu_oldtabindex="1" jax="CHTML" role="application" sre-explorer- style="font-size: 103%; position: relative;" tabindex="0"><mjx-math aria-hidden="true" location="graphic/wrcr70795-math-0002.png"><mjx-semantics><mjx-mrow><mjx-mi data-semantic-annotation="clearspeak:simple" data-semantic-font="italic" data-semantic- data-semantic-role="latinletter" data-semantic-speech="upper N" data-semantic-type="identifier"><mjx-c></mjx-c></mjx-mi></mjx-mrow></mjx-semantics></mjx-math><mjx-assistive-mml display="inline" unselectable="on"><math altimg="urn:x-wiley:00431397:media:wrcr70795:wrcr70795-math-0002" display="inline" location="graphic/wrcr70795-math-0002.png" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi data-semantic-="" data-semantic-annotation="clearspeak:simple" data-semantic-font="ita
{"title":"Regionalization of Optimal Baseflow Separation Using Catchment-Scale Characteristics","authors":"Yongen Lin, Yiwen Mei, Dagang Wang, Jinxin Zhu, Huan Wu, Shuo Wang, Gaohong Yin, Liang Gao, Emmanouil N. Anagnoustou","doi":"10.1029/2025wr040578","DOIUrl":"https://doi.org/10.1029/2025wr040578","url":null,"abstract":"Empirical baseflow filters are widely used for baseflow separation. These filters rely on ad-hoc parameters that introduce significant uncertainties in the calculation. A recent study by Mei et al. (2024, https://doi.org/10.1029/2023wr036386) optimized these parameters using environmental tracer data for 1,100 catchments across the Contiguous United States (CONUS). However, optimized parameters are unavailable for most CONUS catchments lacking tracer data. To address this gap, we developed regionalization models for the filter parameters using the random forest (RF) algorithm and 82 catchment-scale characteristics, including geomorphology, climate, soil properties, and human activities. We demonstrated this approach for the block length parameter <span data-altimg=\"/cms/asset/f43cb71a-820f-4b8b-a8c0-c09af9c296ff/wrcr70795-math-0001.png\"></span><mjx-container ctxtmenu_counter=\"436\" ctxtmenu_oldtabindex=\"1\" jax=\"CHTML\" role=\"application\" sre-explorer- style=\"font-size: 103%; position: relative;\" tabindex=\"0\"><mjx-math aria-hidden=\"true\" location=\"graphic/wrcr70795-math-0001.png\"><mjx-semantics><mjx-mrow><mjx-mi data-semantic-annotation=\"clearspeak:simple\" data-semantic-font=\"italic\" data-semantic- data-semantic-role=\"latinletter\" data-semantic-speech=\"upper N\" data-semantic-type=\"identifier\"><mjx-c></mjx-c></mjx-mi></mjx-mrow></mjx-semantics></mjx-math><mjx-assistive-mml display=\"inline\" unselectable=\"on\"><math altimg=\"urn:x-wiley:00431397:media:wrcr70795:wrcr70795-math-0001\" display=\"inline\" location=\"graphic/wrcr70795-math-0001.png\" xmlns=\"http://www.w3.org/1998/Math/MathML\"><semantics><mrow><mi data-semantic-=\"\" data-semantic-annotation=\"clearspeak:simple\" data-semantic-font=\"italic\" data-semantic-role=\"latinletter\" data-semantic-speech=\"upper N\" data-semantic-type=\"identifier\">N</mi></mrow>$N$</annotation></semantics></math></mjx-assistive-mml></mjx-container> of the smooth minima baseflow filter, one of the optimized filters in Mei et al.’s study, across 855 catchments. Our results show that the prediction of <span data-altimg=\"/cms/asset/816eb4e8-afca-4152-a4ac-77f24d01f79e/wrcr70795-math-0002.png\"></span><mjx-container ctxtmenu_counter=\"437\" ctxtmenu_oldtabindex=\"1\" jax=\"CHTML\" role=\"application\" sre-explorer- style=\"font-size: 103%; position: relative;\" tabindex=\"0\"><mjx-math aria-hidden=\"true\" location=\"graphic/wrcr70795-math-0002.png\"><mjx-semantics><mjx-mrow><mjx-mi data-semantic-annotation=\"clearspeak:simple\" data-semantic-font=\"italic\" data-semantic- data-semantic-role=\"latinletter\" data-semantic-speech=\"upper N\" data-semantic-type=\"identifier\"><mjx-c></mjx-c></mjx-mi></mjx-mrow></mjx-semantics></mjx-math><mjx-assistive-mml display=\"inline\" unselectable=\"on\"><math altimg=\"urn:x-wiley:00431397:media:wrcr70795:wrcr70795-math-0002\" display=\"inline\" location=\"graphic/wrcr70795-math-0002.png\" xmlns=\"http://www.w3.org/1998/Math/MathML\"><semantics><mrow><mi data-semantic-=\"\" data-semantic-annotation=\"clearspeak:simple\" data-semantic-font=\"ita","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"231 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2026-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147478175","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
E. B. Farquhar, E. B. Goldstein, P. J. Bresnahan, B. Settin, J. Stasiewicz, K. Anarde
Chronic flooding is an issue for low-lying coastal communities globally, and it is expected to worsen with rising sea levels. Predicting when and where these floods occur can be difficult as they can be hyper-local and ephemeral, depending on the flood drivers (e.g., tides, rain). These factors make it difficult to measure the full spatial and temporal extent of chronic floods with in situ sensors. Here, we introduce a low-cost (<$400 USD), privacy-preserving camera system that identifies flooding over block-by-block spatial extents at high frequencies (20 s–6 min). Our device—a Tiny Camera with machine learning (ML) (TinyCamML)—is a small, solar-powered, microcontroller-based camera that uses on-device ML to classify images of roadways as containing a “flood” or “no flood.” TinyCamMLs transmit only the classifications (a 1 or 0) to a website in real time, providing situation awareness during flood events over the entire image area while keeping data-transmission costs low and preserving privacy. We demonstrate the TinyCamML's utility during both tidal and compound flood events in North Carolina, USA, which showed differences in flood spatial extents. During this deployment, the TinyCamML detected floods with an 81% accuracy, a 72% precision, and a 90% recall. The utility of the device extends beyond roadway flooding, as the onboard ML model can be easily retrained to capture other rare or ephemeral phenomena.
{"title":"Detection of Coastal Flooding With TinyCamML: A Low-Cost, Privacy-Preserving Cellular-Connected Camera With Onboard ML","authors":"E. B. Farquhar, E. B. Goldstein, P. J. Bresnahan, B. Settin, J. Stasiewicz, K. Anarde","doi":"10.1029/2025wr042023","DOIUrl":"https://doi.org/10.1029/2025wr042023","url":null,"abstract":"Chronic flooding is an issue for low-lying coastal communities globally, and it is expected to worsen with rising sea levels. Predicting when and where these floods occur can be difficult as they can be hyper-local and ephemeral, depending on the flood drivers (e.g., tides, rain). These factors make it difficult to measure the full spatial and temporal extent of chronic floods with in situ sensors. Here, we introduce a low-cost (<$400 USD), privacy-preserving camera system that identifies flooding over block-by-block spatial extents at high frequencies (20 s–6 min). Our device—a Tiny Camera with machine learning (ML) (TinyCamML)—is a small, solar-powered, microcontroller-based camera that uses on-device ML to classify images of roadways as containing a “flood” or “no flood.” TinyCamMLs transmit only the classifications (a 1 or 0) to a website in real time, providing situation awareness during flood events over the entire image area while keeping data-transmission costs low and preserving privacy. We demonstrate the TinyCamML's utility during both tidal and compound flood events in North Carolina, USA, which showed differences in flood spatial extents. During this deployment, the TinyCamML detected floods with an 81% accuracy, a 72% precision, and a 90% recall. The utility of the device extends beyond roadway flooding, as the onboard ML model can be easily retrained to capture other rare or ephemeral phenomena.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"12 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2026-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147478174","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Chenchen Tong, Ching-Sheng Huang, Lloyd R. Townley, Chen Wang
Existing analytical models for permeable reactive barriers (PRBs) treat an upgradient aquifer formation with contaminant sources as either a Dirichlet or Robin boundary condition specified at the upgradient face of the PRB. This study develops two new analytical models for three-dimensional transport in a confined aquifer with a PRB. One model applies governing equations to the upgradient and downgradient formations and PRB in between. The other simplified model represents the PRB as a new adsorptive-reactive Robin matching condition including two terms with a coefficient reflecting the effects of adsorption and reaction in the PRB. Analytical solutions satisfying these models are obtained. Results show the coefficient equals PRB's retardation factor times its thickness divided by the average linear flow velocity. The analytical solutions for dimensionless concentration agree within 5% error under quantitative conditions. Treating an upgradient formation as a Dirichlet boundary condition at PRB's upgradient face is preferable to treating the formation as a Robin boundary condition. A solution depending on the Dirichlet boundary condition only requires measurements of concentration at the boundary for arbitrary values of Peclet number defined as the ratio of PRB's thickness to its longitudinal dispersivity. A solution relying on the Robin boundary condition, however, requires measurements of both concentration and concentration gradient when Peclet number is small, and is applicable for large Peclet number when only concentration measurements are available. A handy tool for designing PRB size is provided.
{"title":"Two New Analytical Models for Three-Dimensional Transport in a Confined Aquifer With a Permeable Reactive Barrier: A New Adsorptive-Reactive Robin Matching Condition","authors":"Chenchen Tong, Ching-Sheng Huang, Lloyd R. Townley, Chen Wang","doi":"10.1029/2025wr041717","DOIUrl":"https://doi.org/10.1029/2025wr041717","url":null,"abstract":"Existing analytical models for permeable reactive barriers (PRBs) treat an upgradient aquifer formation with contaminant sources as either a Dirichlet or Robin boundary condition specified at the upgradient face of the PRB. This study develops two new analytical models for three-dimensional transport in a confined aquifer with a PRB. One model applies governing equations to the upgradient and downgradient formations and PRB in between. The other simplified model represents the PRB as a new adsorptive-reactive Robin matching condition including two terms with a coefficient reflecting the effects of adsorption and reaction in the PRB. Analytical solutions satisfying these models are obtained. Results show the coefficient equals PRB's retardation factor times its thickness divided by the average linear flow velocity. The analytical solutions for dimensionless concentration agree within 5% error under quantitative conditions. Treating an upgradient formation as a Dirichlet boundary condition at PRB's upgradient face is preferable to treating the formation as a Robin boundary condition. A solution depending on the Dirichlet boundary condition only requires measurements of concentration at the boundary for arbitrary values of Peclet number defined as the ratio of PRB's thickness to its longitudinal dispersivity. A solution relying on the Robin boundary condition, however, requires measurements of both concentration and concentration gradient when Peclet number is small, and is applicable for large Peclet number when only concentration measurements are available. A handy tool for designing PRB size is provided.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"12 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2026-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147489478","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Climate change is expected to increase storm-surge intensity while reducing its frequency, posing complex challenges for the recovery of coastal aquifers subject to recurrent wave overwash events. This study quantifies the combined effects of these opposing trends using surface–subsurface integrated numerical simulations of a generalized island aquifer across 12 scenarios with varying storm-surge frequency and intensity. Here, we show that two distinct long-term regimes emerge: (a) full recovery, where the aquifer returns to pre-surge conditions if storm intensity and frequency remain below critical thresholds, and (b) shifted equilibrium, characterized by persistent salt accumulation and depleted fresh groundwater availability if these thresholds are exceeded. Higher hydraulic conductivity and smaller island width exacerbate salt accumulation, the former by increasing the salt load introduced during each storm-surge event, and the latter by decelerating subsequent flushing. The transition between recovery and shifted-equilibrium regimes can be represented with a dimensionless number, <span data-altimg="/cms/asset/a1002d90-5049-44c6-85e3-c647aa0c9571/wrcr70775-math-0001.png"></span><mjx-container ctxtmenu_counter="365" ctxtmenu_oldtabindex="1" jax="CHTML" role="application" sre-explorer- style="font-size: 103%; position: relative;" tabindex="0"><mjx-math aria-hidden="true" location="graphic/wrcr70775-math-0001.png"><mjx-semantics><mjx-mrow><mjx-mi data-semantic-annotation="clearspeak:simple" data-semantic-font="italic" data-semantic- data-semantic-role="latinletter" data-semantic-speech="upper E" data-semantic-type="identifier"><mjx-c></mjx-c></mjx-mi></mjx-mrow></mjx-semantics></mjx-math><mjx-assistive-mml display="inline" unselectable="on"><math altimg="urn:x-wiley:00431397:media:wrcr70775:wrcr70775-math-0001" display="inline" location="graphic/wrcr70775-math-0001.png" xmlns="http://www.w3.org/1998/Math/MathML"><semantics><mrow><mi data-semantic-="" data-semantic-annotation="clearspeak:simple" data-semantic-font="italic" data-semantic-role="latinletter" data-semantic-speech="upper E" data-semantic-type="identifier">E</mi></mrow>$E$</annotation></semantics></math></mjx-assistive-mml></mjx-container>, that integrates the effect of storm-surge intensity and frequency on salt load. In a shifted equilibrium regime, the excess salt load at new dynamic equilibria is effectively approximated by linear functions of <span data-altimg="/cms/asset/07fa268d-c473-4517-9582-851da1a5b62f/wrcr70775-math-0002.png"></span><mjx-container ctxtmenu_counter="366" ctxtmenu_oldtabindex="1" jax="CHTML" role="application" sre-explorer- style="font-size: 103%; position: relative;" tabindex="0"><mjx-math aria-hidden="true" location="graphic/wrcr70775-math-0002.png"><mjx-semantics><mjx-mrow><mjx-mi data-semantic-annotation="clearspeak:simple" data-semantic-font="italic" data-semantic- data-semantic-role="latinletter" data-semantic-speech="upper E" data-semantic-type="identi
{"title":"Climate Change Alters Post-Surge Recovery of Coastal Aquifers","authors":"Satoshi Tajima, René Therrien, Philip Brunner","doi":"10.1029/2025wr042142","DOIUrl":"https://doi.org/10.1029/2025wr042142","url":null,"abstract":"Climate change is expected to increase storm-surge intensity while reducing its frequency, posing complex challenges for the recovery of coastal aquifers subject to recurrent wave overwash events. This study quantifies the combined effects of these opposing trends using surface–subsurface integrated numerical simulations of a generalized island aquifer across 12 scenarios with varying storm-surge frequency and intensity. Here, we show that two distinct long-term regimes emerge: (a) full recovery, where the aquifer returns to pre-surge conditions if storm intensity and frequency remain below critical thresholds, and (b) shifted equilibrium, characterized by persistent salt accumulation and depleted fresh groundwater availability if these thresholds are exceeded. Higher hydraulic conductivity and smaller island width exacerbate salt accumulation, the former by increasing the salt load introduced during each storm-surge event, and the latter by decelerating subsequent flushing. The transition between recovery and shifted-equilibrium regimes can be represented with a dimensionless number, <span data-altimg=\"/cms/asset/a1002d90-5049-44c6-85e3-c647aa0c9571/wrcr70775-math-0001.png\"></span><mjx-container ctxtmenu_counter=\"365\" ctxtmenu_oldtabindex=\"1\" jax=\"CHTML\" role=\"application\" sre-explorer- style=\"font-size: 103%; position: relative;\" tabindex=\"0\"><mjx-math aria-hidden=\"true\" location=\"graphic/wrcr70775-math-0001.png\"><mjx-semantics><mjx-mrow><mjx-mi data-semantic-annotation=\"clearspeak:simple\" data-semantic-font=\"italic\" data-semantic- data-semantic-role=\"latinletter\" data-semantic-speech=\"upper E\" data-semantic-type=\"identifier\"><mjx-c></mjx-c></mjx-mi></mjx-mrow></mjx-semantics></mjx-math><mjx-assistive-mml display=\"inline\" unselectable=\"on\"><math altimg=\"urn:x-wiley:00431397:media:wrcr70775:wrcr70775-math-0001\" display=\"inline\" location=\"graphic/wrcr70775-math-0001.png\" xmlns=\"http://www.w3.org/1998/Math/MathML\"><semantics><mrow><mi data-semantic-=\"\" data-semantic-annotation=\"clearspeak:simple\" data-semantic-font=\"italic\" data-semantic-role=\"latinletter\" data-semantic-speech=\"upper E\" data-semantic-type=\"identifier\">E</mi></mrow>$E$</annotation></semantics></math></mjx-assistive-mml></mjx-container>, that integrates the effect of storm-surge intensity and frequency on salt load. In a shifted equilibrium regime, the excess salt load at new dynamic equilibria is effectively approximated by linear functions of <span data-altimg=\"/cms/asset/07fa268d-c473-4517-9582-851da1a5b62f/wrcr70775-math-0002.png\"></span><mjx-container ctxtmenu_counter=\"366\" ctxtmenu_oldtabindex=\"1\" jax=\"CHTML\" role=\"application\" sre-explorer- style=\"font-size: 103%; position: relative;\" tabindex=\"0\"><mjx-math aria-hidden=\"true\" location=\"graphic/wrcr70775-math-0002.png\"><mjx-semantics><mjx-mrow><mjx-mi data-semantic-annotation=\"clearspeak:simple\" data-semantic-font=\"italic\" data-semantic- data-semantic-role=\"latinletter\" data-semantic-speech=\"upper E\" data-semantic-type=\"identi","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"9 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2026-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147440359","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Helena Bergstedt, Benjamin M. Jones, Mikhail Kanveskiy, Andrew Parsekian, Annett Bartsch, Rustam Khairullin, Chiara Gruber
Arctic lake, bay and lagoon ice regimes play a crucial role in understanding permafrost stability, hydrology, and carbon cycling in permafrost regions. This study integrates field and Synthetic Aperture Radar (SAR) data from Sentinel-1 to improve the classification of Arctic lake, bay and lagoon ice regimes, with a particular focus on the influence of under-ice water salinity. A May 2024 field campaign, revealed that waterbodies with a specific conductance of >4,000 µS/cm at the end of winter were often misclassified as bedfast using traditional SAR-based classification methods. These findings underscore the limitations of relying on established approaches to differentiate floating and bedfast ice and highlight the need to account for differences in under-ice water chemistry. To address this, we developed a novel ice regime classification, incorporating both early and late winter season SAR imagery, early summer SAR imagery to determine ice-off timing, and a digital surface model. This enabled us to reliably classify waterbodies that are influenced by elevated under-ice water salinities. Our approach identifies several floating ice waterbodies with moderate to high specific conductance values that were previously thought to freeze to their bed each winter. This study demonstrates that changes in radar backscatter can result from salinity-related variations in ice and water properties, in addition to waterbody depth and ice thickness. These findings have significant implications for hydrological modeling, ecological studies, winter water availability, and community safety in the Arctic, emphasizing the need for integrating field observations with remote sensing data to better understand ice dynamics in permafrost regions.
{"title":"Improving SAR-Based Classification of Arctic Lake, Bay and Lagoon Ice by Accounting for Under Ice Water Salinity","authors":"Helena Bergstedt, Benjamin M. Jones, Mikhail Kanveskiy, Andrew Parsekian, Annett Bartsch, Rustam Khairullin, Chiara Gruber","doi":"10.1029/2025wr040504","DOIUrl":"https://doi.org/10.1029/2025wr040504","url":null,"abstract":"Arctic lake, bay and lagoon ice regimes play a crucial role in understanding permafrost stability, hydrology, and carbon cycling in permafrost regions. This study integrates field and Synthetic Aperture Radar (SAR) data from Sentinel-1 to improve the classification of Arctic lake, bay and lagoon ice regimes, with a particular focus on the influence of under-ice water salinity. A May 2024 field campaign, revealed that waterbodies with a specific conductance of >4,000 µS/cm at the end of winter were often misclassified as bedfast using traditional SAR-based classification methods. These findings underscore the limitations of relying on established approaches to differentiate floating and bedfast ice and highlight the need to account for differences in under-ice water chemistry. To address this, we developed a novel ice regime classification, incorporating both early and late winter season SAR imagery, early summer SAR imagery to determine ice-off timing, and a digital surface model. This enabled us to reliably classify waterbodies that are influenced by elevated under-ice water salinities. Our approach identifies several floating ice waterbodies with moderate to high specific conductance values that were previously thought to freeze to their bed each winter. This study demonstrates that changes in radar backscatter can result from salinity-related variations in ice and water properties, in addition to waterbody depth and ice thickness. These findings have significant implications for hydrological modeling, ecological studies, winter water availability, and community safety in the Arctic, emphasizing the need for integrating field observations with remote sensing data to better understand ice dynamics in permafrost regions.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"74 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2026-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147439838","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Yuhao Wang, Ke Zhang, Edward Park, Jie Liu, Yuning Luo, Shunzhang Li, Sheng Wang
Understanding how catchments respond to environmental changes is critical for water resource management. However, few studies have systematically linked catchment characteristics, environmental changes, and hydrological responses. Therefore, this study proposes a novel dual-clustering approach for identifying hydrological response patterns. It constructs the catchment characteristic indicator system for the baseline period and introduces dynamic similarity indicators that reflect climate change and anthropogenic impacts to achieve dual clustering, thereby identifying hydrological response differences. Furthermore, it employs the eXtreme Gradient Boosting (XGBoost) and SHapley Additive exPlanations (SHAP) methods to identify key influencing factors of runoff change significance, providing interpretable insights into differences in hydrological responses. The approach is applied to the Haihe River Basin, indicating that 160 catchments are classified into nine static groups (A1–A9) based on catchment characteristics and five dynamic groups (B1–B5) based on environmental changes. The XGBoost model demonstrates good performance in identifying hydrological response patterns, SHAP analysis identifies the top four important factors as percentage of areas with substantial declines in the water table (positive), proportion of natural land use (positive), degree of humidity (negative), and mean elevation (positive). Catchments located in the northwestern mountainous areas are more susceptible to environmental changes, while those located in the southwestern mountainous areas and the southern plains show relatively stable response patterns. Additionally, environmental change patterns characterized by substantial water table decline are more likely to trigger significant runoff change. This approach provides new insights into the effects of interactions between static catchment characteristics and dynamic environmental changes on hydrological functioning.
{"title":"A Novel Dual-Clustering Approach for Identifying Hydrological Response Patterns From Catchment Characteristics and Environmental Changes","authors":"Yuhao Wang, Ke Zhang, Edward Park, Jie Liu, Yuning Luo, Shunzhang Li, Sheng Wang","doi":"10.1029/2025wr041613","DOIUrl":"https://doi.org/10.1029/2025wr041613","url":null,"abstract":"Understanding how catchments respond to environmental changes is critical for water resource management. However, few studies have systematically linked catchment characteristics, environmental changes, and hydrological responses. Therefore, this study proposes a novel dual-clustering approach for identifying hydrological response patterns. It constructs the catchment characteristic indicator system for the baseline period and introduces dynamic similarity indicators that reflect climate change and anthropogenic impacts to achieve dual clustering, thereby identifying hydrological response differences. Furthermore, it employs the eXtreme Gradient Boosting (XGBoost) and SHapley Additive exPlanations (SHAP) methods to identify key influencing factors of runoff change significance, providing interpretable insights into differences in hydrological responses. The approach is applied to the Haihe River Basin, indicating that 160 catchments are classified into nine static groups (A1–A9) based on catchment characteristics and five dynamic groups (B1–B5) based on environmental changes. The XGBoost model demonstrates good performance in identifying hydrological response patterns, SHAP analysis identifies the top four important factors as percentage of areas with substantial declines in the water table (positive), proportion of natural land use (positive), degree of humidity (negative), and mean elevation (positive). Catchments located in the northwestern mountainous areas are more susceptible to environmental changes, while those located in the southwestern mountainous areas and the southern plains show relatively stable response patterns. Additionally, environmental change patterns characterized by substantial water table decline are more likely to trigger significant runoff change. This approach provides new insights into the effects of interactions between static catchment characteristics and dynamic environmental changes on hydrological functioning.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"8 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2026-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147383879","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：