Recent studies show variations in precipitation-gridded data set accuracy with changing geographical parameters. Ensemble precipitation products, combining diverse data sets, offer global-scale effectiveness, but applying them to regional studies, particularly in small to medium-sized sub-basins, presents challenges in addressing precipitation dependence on specific geographical conditions. Here, we present a newly developed Clusters Based-Minimum Error approach to assimilate different open-source gridded precipitation data sets for forming an accurate precipitation product over small to medium-sized hilly terrain basins, with limited precipitation gauges. This methodology generates the New Gridded Precipitation Data Set (NGPD) from 1991 to 2022 for the Upper Ganga Basin in the western Himalaya, covering approximately 22,292 km2. The study utilizes nine open-source gridded precipitation data sets and 11 observed precipitation gauges, NGPD is evaluated through station-wise, grid-wise, and elevation-wise analyses using statistical parameters, quantile-quantile plots, daily coefficient of determination, Rainfall Anomaly Index, and seasonality/precipitation pattern analyses. Results demonstrate the superior performance of NGPD compared to other gridded precipitation sources across various evaluation metrics. Nash-Sutcliffe Efficiency (NSE), Coefficient of determination (R2), and Root mean squared error (RMSE) range from 0.67 to 0.90, 0.73–0.93, and 4.4–10.69 mm/day, respectively, w.r.t 11 observed precipitation gauges. NGPD outperforms the widely used IMD data set in India, exhibiting a monthly scale improvement of 18.47% and 17.7% in average NSE and R2 values, respectively. Additionally, the methodology is also successfully applied to the Tamor Basin in Nepal, proving its reliability for various Himalayan regions. This approach reliably creates accurate gridded precipitation data sets for hilly sub-basins, especially in Himalayan regions with limited station data.
{"title":"A Cluster-Based Data Assimilation Approach to Generate New Daily Gridded Time Series Precipitation Data in the Himalayan River Basins","authors":"Japjeet Singh, Vishal Singh, Chandra Shekhar Prasad Ojha","doi":"10.1029/2024wr037324","DOIUrl":"https://doi.org/10.1029/2024wr037324","url":null,"abstract":"Recent studies show variations in precipitation-gridded data set accuracy with changing geographical parameters. Ensemble precipitation products, combining diverse data sets, offer global-scale effectiveness, but applying them to regional studies, particularly in small to medium-sized sub-basins, presents challenges in addressing precipitation dependence on specific geographical conditions. Here, we present a newly developed Clusters Based-Minimum Error approach to assimilate different open-source gridded precipitation data sets for forming an accurate precipitation product over small to medium-sized hilly terrain basins, with limited precipitation gauges. This methodology generates the New Gridded Precipitation Data Set (NGPD) from 1991 to 2022 for the Upper Ganga Basin in the western Himalaya, covering approximately 22,292 km<sup>2</sup>. The study utilizes nine open-source gridded precipitation data sets and 11 observed precipitation gauges, NGPD is evaluated through station-wise, grid-wise, and elevation-wise analyses using statistical parameters, quantile-quantile plots, daily coefficient of determination, Rainfall Anomaly Index, and seasonality/precipitation pattern analyses. Results demonstrate the superior performance of NGPD compared to other gridded precipitation sources across various evaluation metrics. Nash-Sutcliffe Efficiency (NSE), Coefficient of determination (<i>R</i><sup>2</sup>), and Root mean squared error (RMSE) range from 0.67 to 0.90, 0.73–0.93, and 4.4–10.69 mm/day, respectively, w.r.t 11 observed precipitation gauges. NGPD outperforms the widely used IMD data set in India, exhibiting a monthly scale improvement of 18.47% and 17.7% in average NSE and <i>R</i><sup>2</sup> values, respectively. Additionally, the methodology is also successfully applied to the Tamor Basin in Nepal, proving its reliability for various Himalayan regions. This approach reliably creates accurate gridded precipitation data sets for hilly sub-basins, especially in Himalayan regions with limited station data.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"122 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2025-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142990590","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}
José Cunha Teixeira, Ludovic Bodet, Agnès Rivière, Amélie Hallier, Alexandrine Gesret, Marine Dangeard, Amine Dhemaied, Joséphine Boisson Gaboriau
Monitoring groundwater tables (GWTs) remains challenging due to limited spatial and temporal observations. This study introduces an innovative approach combining an artificial neural network, specifically a multilayer perceptron (MLP), with continuous passive Multichannel Analysis of Surface Waves (passive-MASW) to construct GWT depth maps. The geologically well-constrained study site includes two piezometers and a permanent 2D geophone array recording train-induced surface waves. At each point of the array, dispersion curves (DCs), displaying Rayleigh-wave phase velocities over a frequency range of 5–50 Hz, were measured daily from December 2022 to September 2023, and latter resampled over wavelengths from 4 to 15 m, to focus on the expected GWT depths (1–5 m). Nine months of daily data near one piezometer, spanning both low and high water periods, were used to train the MLP model. GWT depths were then estimated across the geophone array, producing daily GWT maps. The model's performance was evaluated by comparing inferred GWT depths with observed measurements at the second piezometer. Results show a coefficient of determination (R2) of 80% at the training piezometer and of 68% at the test piezometer, and a remarkably low root-mean-square error (RMSE) of 0.03 m at both locations. These findings highlight the potential of deep learning to estimate GWT maps from seismic data with spatially limited piezometric information, offering a practical and efficient solution for monitoring groundwater dynamics across large spatial extents.
{"title":"Physics-Guided Deep Learning Model for Daily Groundwater Table Maps Estimation Using Passive Surface-Wave Dispersion","authors":"José Cunha Teixeira, Ludovic Bodet, Agnès Rivière, Amélie Hallier, Alexandrine Gesret, Marine Dangeard, Amine Dhemaied, Joséphine Boisson Gaboriau","doi":"10.1029/2024wr037706","DOIUrl":"https://doi.org/10.1029/2024wr037706","url":null,"abstract":"Monitoring groundwater tables (GWTs) remains challenging due to limited spatial and temporal observations. This study introduces an innovative approach combining an artificial neural network, specifically a multilayer perceptron (MLP), with continuous passive Multichannel Analysis of Surface Waves (passive-MASW) to construct GWT depth maps. The geologically well-constrained study site includes two piezometers and a permanent 2D geophone array recording train-induced surface waves. At each point of the array, dispersion curves (DCs), displaying Rayleigh-wave phase velocities <span data-altimg=\"/cms/asset/f0d9b693-7331-4aa5-a654-78f5ac9fba29/wrcr27656-math-0001.png\"></span><math altimg=\"urn:x-wiley:00431397:media:wrcr27656:wrcr27656-math-0001\" display=\"inline\" location=\"graphic/wrcr27656-math-0001.png\">\u0000<semantics>\u0000<mrow>\u0000<mfenced close=\")\" open=\"(\" separators=\"\">\u0000<msub>\u0000<mi>V</mi>\u0000<mi>R</mi>\u0000</msub>\u0000</mfenced>\u0000</mrow>\u0000$left({V}_{R}right)$</annotation>\u0000</semantics></math> over a frequency range of 5–50 Hz, were measured daily from December 2022 to September 2023, and latter resampled over wavelengths from 4 to 15 m, to focus on the expected GWT depths (1–5 m). Nine months of daily <span data-altimg=\"/cms/asset/0f52fe16-966b-40c5-b835-263b40d7b16f/wrcr27656-math-0002.png\"></span><math altimg=\"urn:x-wiley:00431397:media:wrcr27656:wrcr27656-math-0002\" display=\"inline\" location=\"graphic/wrcr27656-math-0002.png\">\u0000<semantics>\u0000<mrow>\u0000<msub>\u0000<mi>V</mi>\u0000<mi>R</mi>\u0000</msub>\u0000</mrow>\u0000${V}_{R}$</annotation>\u0000</semantics></math> data near one piezometer, spanning both low and high water periods, were used to train the MLP model. GWT depths were then estimated across the geophone array, producing daily GWT maps. The model's performance was evaluated by comparing inferred GWT depths with observed measurements at the second piezometer. Results show a coefficient of determination (R<sup>2</sup>) of 80% at the training piezometer and of 68% at the test piezometer, and a remarkably low root-mean-square error (RMSE) of 0.03 m at both locations. These findings highlight the potential of deep learning to estimate GWT maps from seismic data with spatially limited piezometric information, offering a practical and efficient solution for monitoring groundwater dynamics across large spatial extents.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"18 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2025-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142991576","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}
The reliable prediction of sediment transport capacity (Tc) is essential for soil erosion models. Although rock fragments are a common surface cover type, quantitative studies on their relationship with Tc are limited. Tc typically follows a power function with slope gradient (S) and flow discharge (q) under bare flumes, but varying exponents complicate practical application. This study aims to investigate the effect of rock fragment cover on Tc, explore the interactive effects of S, q, and cover on Tc, and ultimately develop a universal Tc prediction equation and assess its feasibility for different scenarios. Flume experiments on Tc with rock fragment cover have been conducted, and many existing Tc prediction equations have been reviewed. The results revealed that the effects of S and q on the relationship between rock fragment cover and Tc were minor and that the impact of rock fragment cover on the relationships of S and q with Tc was also not significant. Consequently, a new universal equation for Tc incorporating cover was developed. This equation featured fixed exponents of 1.66 for S and 1.22 for q and was applicable across various slope gradient, flow discharge, coverage and cover type conditions. Moreover, the impact of rock fragment cover on Tc reduction was significantly less than those of litter cover and stem basal cover (P < 0.05). Therefore, the role of rock fragments should be considered separately in soil erosion models. These findings could significantly advance the practical application of the Tc prediction equation.
{"title":"Effects of Rock Fragment Cover on the Sediment Transport Capacity of Overland Flow","authors":"Lixia Dong, Suhua Fu, Baoyuan Liu","doi":"10.1029/2024wr038621","DOIUrl":"https://doi.org/10.1029/2024wr038621","url":null,"abstract":"The reliable prediction of sediment transport capacity (<i>Tc</i>) is essential for soil erosion models. Although rock fragments are a common surface cover type, quantitative studies on their relationship with <i>Tc</i> are limited. <i>Tc</i> typically follows a power function with slope gradient (<i>S</i>) and flow discharge (<i>q</i>) under bare flumes, but varying exponents complicate practical application. This study aims to investigate the effect of rock fragment cover on <i>Tc</i>, explore the interactive effects of <i>S</i>, <i>q</i>, and cover on <i>Tc</i>, and ultimately develop a universal <i>Tc</i> prediction equation and assess its feasibility for different scenarios. Flume experiments on <i>Tc</i> with rock fragment cover have been conducted, and many existing <i>Tc</i> prediction equations have been reviewed. The results revealed that the effects of <i>S</i> and <i>q</i> on the relationship between rock fragment cover and <i>Tc</i> were minor and that the impact of rock fragment cover on the relationships of <i>S</i> and <i>q</i> with <i>Tc</i> was also not significant. Consequently, a new universal equation for <i>Tc</i> incorporating cover was developed. This equation featured fixed exponents of 1.66 for <i>S</i> and 1.22 for <i>q</i> and was applicable across various slope gradient, flow discharge, coverage and cover type conditions. Moreover, the impact of rock fragment cover on <i>Tc</i> reduction was significantly less than those of litter cover and stem basal cover (<i>P</i> < 0.05). Therefore, the role of rock fragments should be considered separately in soil erosion models. These findings could significantly advance the practical application of the <i>Tc</i> prediction equation.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"15 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2025-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142991580","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}
Acoustic Doppler current profilers (ADCPs) are a global standard in observing flow fields in rivers, estuaries and the coastal ocean. To date, it remains a labor intensive challenge to isolate mean flow fields governed by river discharge, tides and atmospheric forcing on the one hand, from small-scale turbulence, positioning imprecision, Doppler noise and erroneous backscatter, on the other hand. Here, we introduce a generic, new method of combining raw shipborne ADCP transect data with continuity and smoothness constraints to obtain better estimates of turbulence-averaged three-dimensional flow velocities in any type of open water body. The physical constraints are enforced with variable relative importance via generalized Tikhonov regularization. We demonstrate that in complex estuarine flow, this procedure allows for more reliable estimates of tidal amplitudes, phases and their gradients than what is possible with a purely data-based approach, by testing the method's generalization capabilities and robustness to turbulence and measurement noise on a data set retrieved at a tidal channel junction. The increased adherence to mass conservation and robustness to noise of various kinds allows for more reliable and verifiable estimates of Reynolds-averaged flow components, and subsequently, of terms in the Navier-Stokes equations.
{"title":"Physics-Informed Estimation of Tidal and Subtidal Flow Fields From ADCP Repeat Transect Data","authors":"H. Jongbloed, B. Vermeulen, A. J. F. Hoitink","doi":"10.1029/2023wr036038","DOIUrl":"https://doi.org/10.1029/2023wr036038","url":null,"abstract":"Acoustic Doppler current profilers (ADCPs) are a global standard in observing flow fields in rivers, estuaries and the coastal ocean. To date, it remains a labor intensive challenge to isolate mean flow fields governed by river discharge, tides and atmospheric forcing on the one hand, from small-scale turbulence, positioning imprecision, Doppler noise and erroneous backscatter, on the other hand. Here, we introduce a generic, new method of combining raw shipborne ADCP transect data with continuity and smoothness constraints to obtain better estimates of turbulence-averaged three-dimensional flow velocities in any type of open water body. The physical constraints are enforced with variable relative importance via generalized Tikhonov regularization. We demonstrate that in complex estuarine flow, this procedure allows for more reliable estimates of tidal amplitudes, phases and their gradients than what is possible with a purely data-based approach, by testing the method's generalization capabilities and robustness to turbulence and measurement noise on a data set retrieved at a tidal channel junction. The increased adherence to mass conservation and robustness to noise of various kinds allows for more reliable and verifiable estimates of Reynolds-averaged flow components, and subsequently, of terms in the Navier-Stokes equations.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"122 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2025-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142991583","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}
Pollutant transport in discrete fracture networks (DFNs) exhibits complex dynamics that challenge reliable model predictions, even with detailed fracture data. To address this issue, this study derives an upscaled integral-differential equation to predict transient anomalous diffusion in two-dimensional (2D) DFNs. The model includes both transmissive and dead-end fractures (DEFs), where stagnant water zones in DEFs cause non-uniform flow and transient sub-diffusive transport, as shown by both literature and DFN flow and transport simulations using COMSOL. The upscaled model's main parameters are quantitatively linked to fracture properties, especially the probability density function of DEF lengths. Numerical experiments show the model's accuracy in predicting the full-term evolution of conservative tracers in 2D DFNs with power-law distributed fracture lengths and two orientation sets. Field applications indicate that while model parameters for transient sub-diffusion can be predicted from observed DFN distributions, predicting parameters controlling solute displacement in transmissive fractures requires additional field work, such as tracer tests. Parameter sensitivity analysis further correlates late-time solute transport dynamics with fracture properties, such as fracture density and average length. Potential extensions of the upscaled model are also discussed. This study, therefore, proves that transient anomalous transport in 2D DFNs with DEFs can be at least partially predicted, offering an initial step toward improving model predictions for pollutant transport in real-world fractured aquifer systems.
{"title":"Predicting Transient Anomalous Transport in Two-Dimensional Discrete Fracture Networks With Dead-End Fractures","authors":"HongGuang Sun, Dawei Lei, Yong Zhang, Jiazhong Qian, Xiangnan Yu","doi":"10.1029/2024wr038731","DOIUrl":"https://doi.org/10.1029/2024wr038731","url":null,"abstract":"Pollutant transport in discrete fracture networks (DFNs) exhibits complex dynamics that challenge reliable model predictions, even with detailed fracture data. To address this issue, this study derives an upscaled integral-differential equation to predict transient anomalous diffusion in two-dimensional (2D) DFNs. The model includes both transmissive and dead-end fractures (DEFs), where stagnant water zones in DEFs cause non-uniform flow and transient sub-diffusive transport, as shown by both literature and DFN flow and transport simulations using COMSOL. The upscaled model's main parameters are quantitatively linked to fracture properties, especially the probability density function of DEF lengths. Numerical experiments show the model's accuracy in predicting the full-term evolution of conservative tracers in 2D DFNs with power-law distributed fracture lengths and two orientation sets. Field applications indicate that while model parameters for transient sub-diffusion can be predicted from observed DFN distributions, predicting parameters controlling solute displacement in transmissive fractures requires additional field work, such as tracer tests. Parameter sensitivity analysis further correlates late-time solute transport dynamics with fracture properties, such as fracture density and average length. Potential extensions of the upscaled model are also discussed. This study, therefore, proves that transient anomalous transport in 2D DFNs with DEFs can be at least partially predicted, offering an initial step toward improving model predictions for pollutant transport in real-world fractured aquifer systems.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"5 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2025-01-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142989104","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}
Ian Gambill, Anna Marshall, David A. Benson, Sawyer McFadden, Alexis Navarre-Sitchler, Ellen Wohl, Kamini Singha
Here, we explore how differences in morphologic heterogeneity due to logjams and secondary channels drive transient storage across discharge in two stream reaches within the Front Range of Colorado, USA. During three tracer tests conducted from baseflow to near-peak snowmelt, we collected instream fluid conductivity measurements and conducted electrical resistivity surveys to characterize tracer movement in the surface and subsurface of the stream system. The reach with two logjams and an intermittent secondary channel exhibited greater heterogeneity in surface transient storage, driving heterogeneity in hyporheic exchange flows, compared to the reach with a single logjam and a perennial secondary channel. As discharge increased, (a) backwater pools created by logjams increased in size in both systems, (b) channel complexity increased as logjams forced flow into secondary channels, and (c) subsurface flowpath distribution increased. Various transient storage indices provide some insight on solute retention but compressing data from this system into simple values was unintuitive given the noise in breakthrough-curve tails and secondary peaks in concentration. While subsurface exchange increases with discharge in both reaches, retention may not. Flushing of subsurface tracers is highest at medium discharge as interpreted from the electrical resistivity inversions in both reaches, perhaps because of a tradeoff between the increasing extent of subsurface flowpaths with discharge and larger pressure gradients for driving flow. This work is one of the first to explore controls on exchange and retention in stream systems with multiple logjams and evolving channel planform using geophysical data to constrain the subsurface movement of solutes.
{"title":"Exploring the Influence of Morphologic Heterogeneity and Discharge on Transient Storage in Stream Systems: 1. Insights From the Field","authors":"Ian Gambill, Anna Marshall, David A. Benson, Sawyer McFadden, Alexis Navarre-Sitchler, Ellen Wohl, Kamini Singha","doi":"10.1029/2023wr036031","DOIUrl":"https://doi.org/10.1029/2023wr036031","url":null,"abstract":"Here, we explore how differences in morphologic heterogeneity due to logjams and secondary channels drive transient storage across discharge in two stream reaches within the Front Range of Colorado, USA. During three tracer tests conducted from baseflow to near-peak snowmelt, we collected instream fluid conductivity measurements and conducted electrical resistivity surveys to characterize tracer movement in the surface and subsurface of the stream system. The reach with two logjams and an intermittent secondary channel exhibited greater heterogeneity in surface transient storage, driving heterogeneity in hyporheic exchange flows, compared to the reach with a single logjam and a perennial secondary channel. As discharge increased, (a) backwater pools created by logjams increased in size in both systems, (b) channel complexity increased as logjams forced flow into secondary channels, and (c) subsurface flowpath distribution increased. Various transient storage indices provide some insight on solute retention but compressing data from this system into simple values was unintuitive given the noise in breakthrough-curve tails and secondary peaks in concentration. While subsurface exchange increases with discharge in both reaches, retention may not. Flushing of subsurface tracers is highest at medium discharge as interpreted from the electrical resistivity inversions in both reaches, perhaps because of a tradeoff between the increasing extent of subsurface flowpaths with discharge and larger pressure gradients for driving flow. This work is one of the first to explore controls on exchange and retention in stream systems with multiple logjams and evolving channel planform using geophysical data to constrain the subsurface movement of solutes.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"7 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2025-01-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142989106","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}
Hiren Solanki, Urmin Vegad, Anuj Kushwaha, Vimal Mishra
Streamflow prediction is crucial for flood monitoring and early warning, which often hampered by bias and uncertainties arising from nonlinear processes, model parameterization, and errors in meteorological forecast. We examined the utility of multiple hydrological models (VIC, H08, CWatM, Noah-MP, and CLM) and machine learning (ML) methods to improve streamflow simulations and prediction. The hydrological models (HMs) were forced with observed meteorological data from the India Meteorological Department (IMD) and meteorological forecast from the Global Ensemble Forecast System (GEFS) to simulate flood peaks and flood inundation areas. We used Multiple Linear Regression, Random Forest (RF), Extreme Gradient Boosting (XGB), and Long Short-Term Memory (LSTM) for the post-processing of simulated streamflow from HMs. Considering the influence of dams is crucial for the effectiveness of HMs and ML methods for improving streamflow simulations and predictions. In addition, ML-based multi-model ensemble streamflow from HMs performs better than individual models, highlighting the need for multi-model-based streamflow forecast systems. The post-processing of streamflow simulated by the hydrological models using ML significantly improved overall streamflow simulations, with limited improvement in high-flow conditions. The combination of physics-based hydrological models, observed climate data, and ML methods improve streamflow predictions for flood magnitude, timing, and inundated area, which can be valuable for developing flood early warning systems in India.
{"title":"Improving Streamflow Prediction Using Multiple Hydrological Models and Machine Learning Methods","authors":"Hiren Solanki, Urmin Vegad, Anuj Kushwaha, Vimal Mishra","doi":"10.1029/2024wr038192","DOIUrl":"https://doi.org/10.1029/2024wr038192","url":null,"abstract":"Streamflow prediction is crucial for flood monitoring and early warning, which often hampered by bias and uncertainties arising from nonlinear processes, model parameterization, and errors in meteorological forecast. We examined the utility of multiple hydrological models (VIC, H08, CWatM, Noah-MP, and CLM) and machine learning (ML) methods to improve streamflow simulations and prediction. The hydrological models (HMs) were forced with observed meteorological data from the India Meteorological Department (IMD) and meteorological forecast from the Global Ensemble Forecast System (GEFS) to simulate flood peaks and flood inundation areas. We used Multiple Linear Regression, Random Forest (RF), Extreme Gradient Boosting (XGB), and Long Short-Term Memory (LSTM) for the post-processing of simulated streamflow from HMs. Considering the influence of dams is crucial for the effectiveness of HMs and ML methods for improving streamflow simulations and predictions. In addition, ML-based multi-model ensemble streamflow from HMs performs better than individual models, highlighting the need for multi-model-based streamflow forecast systems. The post-processing of streamflow simulated by the hydrological models using ML significantly improved overall streamflow simulations, with limited improvement in high-flow conditions. The combination of physics-based hydrological models, observed climate data, and ML methods improve streamflow predictions for flood magnitude, timing, and inundated area, which can be valuable for developing flood early warning systems in India.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"15 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2025-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142987986","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}
The tidal behavior of a well in semiconfined aquifers can be described by a diffusion equation including a leakage term. This approach is valid for thin aquifers, as long as the aquitard has low permeability relative to the aquifer. However, in cases where the aquifer is thick and the permeability of the aquitard is not low, using the existing solutions based on these approximations leads to unsatisfactory outcomes. Alternative solutions for both vertical and horizontal wells were obtained by solving the standard diffusion equation, with leakage expressed as a boundary condition. The solutions can be used to estimate any one of wellbore storage coefficient, skin effect, hydraulic diffusivity, and vertical leakage, given the other three. Furthermore, a nondimensional number, named hydraulic Biot number, was derived mathematically, which forms the basis for a quantitative criterion to assess the applicability of existing solutions. In the case of a vertical well, the existing solution exhibits acceptable error only if the hydraulic Biot number is less than 0.245. The new solution extends this upper limitation to 0.475. However, when the number is greater than 0.475, both the existing solution and new solution are invalid due to the invalid uniform flowrate assumption. For a horizontal well, when the number is less than 0.245, the existing solution is suitable with acceptable error. Our new solution effectively overcomes this limitation. Finally, the new solution was applied to the case of the Arbuckle aquifer to demonstrate the improved validity of the new solution compared to the existing one.
{"title":"Analytical Solutions for Groundwater Response to Earth Tides in Thick Semiconfined Aquifers","authors":"Xunfeng Lu, Kozo Sato, Roland N. Horne","doi":"10.1029/2023wr036237","DOIUrl":"https://doi.org/10.1029/2023wr036237","url":null,"abstract":"The tidal behavior of a well in semiconfined aquifers can be described by a diffusion equation including a leakage term. This approach is valid for thin aquifers, as long as the aquitard has low permeability relative to the aquifer. However, in cases where the aquifer is thick and the permeability of the aquitard is not low, using the existing solutions based on these approximations leads to unsatisfactory outcomes. Alternative solutions for both vertical and horizontal wells were obtained by solving the standard diffusion equation, with leakage expressed as a boundary condition. The solutions can be used to estimate any one of wellbore storage coefficient, skin effect, hydraulic diffusivity, and vertical leakage, given the other three. Furthermore, a nondimensional number, named hydraulic Biot number, was derived mathematically, which forms the basis for a quantitative criterion to assess the applicability of existing solutions. In the case of a vertical well, the existing solution exhibits acceptable error only if the hydraulic Biot number is less than 0.245. The new solution extends this upper limitation to 0.475. However, when the number is greater than 0.475, both the existing solution and new solution are invalid due to the invalid uniform flowrate assumption. For a horizontal well, when the number is less than 0.245, the existing solution is suitable with acceptable error. Our new solution effectively overcomes this limitation. Finally, the new solution was applied to the case of the Arbuckle aquifer to demonstrate the improved validity of the new solution compared to the existing one.","PeriodicalId":23799,"journal":{"name":"Water Resources Research","volume":"328 1","pages":""},"PeriodicalIF":5.4,"publicationDate":"2025-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142988070","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：