Advances in Water Resources最新文献

{"title":"Field-scale soil moisture dynamics predicted by deep learning","authors":"Sahar Bakhshian ,&nbsp;Negar Zarepakzad ,&nbsp;Hannes Nevermann ,&nbsp;Cathy Hohenegger ,&nbsp;Dani Or ,&nbsp;Nima Shokri","doi":"10.1016/j.advwatres.2025.104976","DOIUrl":"10.1016/j.advwatres.2025.104976","url":null,"abstract":"<div><div>Soil moisture plays a critical role in land–atmosphere interactions. Prediction of its dynamics is still a grand challenge. While in-situ measurements using sensors offer highly temporally resolved and accurate information compared to satellite observations, existing sensor networks are sparse and scarce. Here we propose a deep learning model for bridging the gap between infrequent satellite observations and sparse in-situ sensor network to improve near-term soil moisture predictions. The Long Short-Term Memory (LSTM)-based deep learning model was used to forecast soil moisture dynamics using soil parameters and climatic variables (e.g. air temperature, relative humidity, pressure, wind speed, turbulent fluxes, solar and terrestrial waves) collected from a dense network of sensors in a field located in Germany in an area of about 20 hectares. The dynamic time-lagged cross-correlation between soil moisture and other co-located soil and climatic features was calculated and a set of optimal predictors for training the LSTM model was selected. To efficiently learn the long-term dependency of soil moisture on its historical trends and to improve the prediction capability of the model, we optimized the LSTM structure, hyperparameters, and the size of the sliding window based on the goodness of fit (<span><math><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup></math></span> score) of the model. We also examined the feasibility of employing the model developed using temporal data from one location for the prediction of soil moisture at other locations across the landscape. The results illustrate the robustness and efficiency of the proposed model for the spatio-temporal prediction of soil moisture. <strong>Plain Language Summary</strong> Understanding and monitoring soil moisture dynamics is crucial affecting ecosystem health, climate and extreme weather patterns, and the agricultural sector. However, predicting the temporal and spatial variation of soil moisture is challenging because of the complex interactions between the land and atmosphere. While soil moisture measurement with in-situ ground-based sensors provide a high level of temporal frequency in comparison to satellite data, the implementation of dense monitoring networks to capture spatial variability of soil moisture is not economically viable. To address this problem, we utilized machine learning techniques to predict temporal and spatial variation of soil moisture using data we measured in a field in Germany. The developed model was examined against the experimental data with the results illustrating that AI-based solutions could offer a powerful tool to predict soil moisture dynamics.</div></div>","PeriodicalId":7614,"journal":{"name":"Advances in Water Resources","volume":"201 ","pages":"Article 104976"},"PeriodicalIF":4.0,"publicationDate":"2025-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143873023","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Discontinuous Galerkin simulator of shallow vortical flow with turbulence","authors":"Georges Kesserwani,&nbsp;Xitong Sun ,&nbsp;Mahya Hajihassanpour ,&nbsp;Mohammad Kazem Sharifian","doi":"10.1016/j.advwatres.2025.104986","DOIUrl":"10.1016/j.advwatres.2025.104986","url":null,"abstract":"<div><div>Shallow vortical flow can often occurs past (un)submerged topographies, prevailing in quasi-steady states with turbulence. Practically, vortical flow is represented by the two-dimensional (2D) Reynolds-Averaged Navier–Stokes equations, including the two-equation <em>k-</em>ε turbulent model (RANS-<em>k-</em>ε), and are commonly resolved by finite difference/volumes second-order accurate solvers. Such RANS-<em>k-</em>ε solvers, in addition to needing a fine resolution, require adding artificial treatments–extrinsic (unlocalised) reconstructions of wet-dry fronts with slope-limiting–that can impact the vortical eddy predictions. The second-order discontinuous Galerkin (DG2) solver intrinsically integrates the wet-dry fronts and uses localised slope-limiting; resulting in an implicit large eddy simulator with the shallow water equations (DG2-SWE) that can only simulate uncompounded eddies. A novel DG2 solver of RANS-<em>k-</em>ε (DG2-RANS-<em>k-</em>ε) is devised for simulating a wider range of vortical eddies, by: first, transforming the 5×5 advective-diffusive RANS-<em>k-</em>ε system into a 13×13 advection-dominated system; second, extending the DG2 formulation to the 13×13 system, with adaptation of its robustness treatments for the mean-flow variables; and, last, adding a new combination of stability/positivity-preserving treatments for turbulent-flow quantities. The DG2-RANS<em>-k-</em>ε solver is evaluated for simulating five experimental benchmarks using coarse, medium and fine resolutions. Results show that DG2-RANS<em>-k-</em>ε can reproduce compound eddies from the medium resolution, and that DG2-RANS (without <em>k-</em>ε) can better reproduce laminar wakes. Using the medium resolution reduces runtimes by 7-fold and running on the GPU further reduce runtimes by 2-to-6-fold. The code, including simulation setup files, is open-source within a new release of the LISFLOOD-FP hydraulic modelling environment (https://doi.org/10.5281/zenodo.7628739), with documentation and demonstration videos (<span><span>https://www.seamlesswave.com/DG2_RANS</span><svg><path></path></svg></span>).</div></div>","PeriodicalId":7614,"journal":{"name":"Advances in Water Resources","volume":"201 ","pages":"Article 104986"},"PeriodicalIF":4.0,"publicationDate":"2025-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143870565","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Effects of temperature on the multi-scale characteristics of bioclogging in porous media","authors":"Shilin Wang,&nbsp;Xixiang Li,&nbsp;Lijian Huang,&nbsp;Gengyang Zang,&nbsp;Taijia Lu,&nbsp;Yanfeng Gong,&nbsp;Liping Chen","doi":"10.1016/j.advwatres.2025.104980","DOIUrl":"10.1016/j.advwatres.2025.104980","url":null,"abstract":"<div><div>In this study, an improved multi-scale algorithm was developed to analyze the effects of temperature on the bioclogging processes at pore and Representative Elementary Volume (REV) scales. In this algorithm, Immersed Boundary-Lattice Boltzmann Method- Cellular Automata (IB-LBM-CA) model for pore-scale simulation and Discrete Unified Gas-Kinetic Scheme-Cellular Automata (DUGKS-CA) model for REV simulation were coupled with the porosity and permeability obtained at pore-scale simulation as connecting bridge. In this study, a biofilm detachment model with non-Newtonian shear-thinning characteristics was considered, and also the tendency of high nutrient concentration for microbes and the inoculation rates of microbes were taken into account. The simulation results had been validated by laboratory percolation experiments with a high consistence. The main results are as follows: (1) At the pore scale, as the temperature increased, the clogging time decreased, the proportion of the clogging time occupied by slow decline period decreased and the proportion of the clogging time taken by rapid decline period increased. (2) At the REV scale, the clogging time decreased with the increasing temperature and inoculation rate. The relationship between clogging time and inoculation rate could be described by an exponential decay model. (3) A linear relationship was found between the clogging time at the pore and REV scales, and the proportion coefficient decreased as the inoculation rate increased. The influences of inoculation rates and temperatures on the clogging time proportion coefficients could be described by exponential decay model and linear decay model, respectively. The average exponential decay rate of clogging time proportion coefficient for the inoculation rate is 0.54 / % and the linear decay rates of clogging time proportion coefficient for the temperature are mainly decreased with increasing inoculation rate.</div></div>","PeriodicalId":7614,"journal":{"name":"Advances in Water Resources","volume":"201 ","pages":"Article 104980"},"PeriodicalIF":4.0,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143838212","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Investigating vertical and lateral gas migration during thermal conduction heating in heterogeneous porous media","authors":"Liam M. Price,&nbsp;Kevin G. Mumford","doi":"10.1016/j.advwatres.2025.104983","DOIUrl":"10.1016/j.advwatres.2025.104983","url":null,"abstract":"<div><div>The successful treatment of contaminated soil and groundwater using thermal remediation technologies relies on the capture and treatment of contaminant vapour produced during heating. The migration of that vapour is affected by subsurface heterogeneity, which must be understood to ensure capture and to prevent condensation outside of a target heating zone. Bench-scale thermal conduction heating experiments were conducted to investigate the migration of steam through homogeneous and heterogeneous porous media. The steam pattern was parabolic adjacent to the heater during homogeneous experiments. In heterogeneous experiments, gas accumulation and migration were observed underneath the capillary barrier. Despite a connected gas pathway to the atmosphere, the transmissivity of the capillary barrier was not sufficient to prevent this migration. This has implications for the transport of heat and contaminant vapour outside of the heated zone, emphasizing the need for effective vapour capture.</div></div>","PeriodicalId":7614,"journal":{"name":"Advances in Water Resources","volume":"201 ","pages":"Article 104983"},"PeriodicalIF":4.0,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143874059","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Tackling water table depth modeling via machine learning: From proxy observations to verifiability","authors":"Joseph Janssen,&nbsp;Ardalan Tootchi,&nbsp;Ali A. Ameli","doi":"10.1016/j.advwatres.2025.104955","DOIUrl":"10.1016/j.advwatres.2025.104955","url":null,"abstract":"<div><div>Spatial patterns of water table depth (WTD) play a crucial role in shaping ecological resilience, hydrological connectivity, and human-centric systems. Generally, a large-scale (e.g., continental or global) continuous map of static WTD can be simulated using either physically-based (PB) or machine learning-based (ML) models. We construct three fine-resolution (500 m) ML simulations of WTD, using the XGBoost algorithm and more than 20 million real and proxy observations of WTD, across the United States and Canada. The three ML models were constrained using known physical relations between WTD’s drivers and WTD and were trained by sequentially adding real and proxy observations of WTD. Through an extensive (pixel-by-pixel) evaluation across the study region and within ten major ecoregions of North America, we demonstrate that our models (corr = 0.6-0.75) can more accurately predict unseen real and proxy observations of WTD compared to two available PB simulations of WTD (corr = 0.21-0.40). However, we still argue that currently-available large-scale simulations of static WTD could be uncertain within data-scarce regions such as steep mountainous regions. We reason that biased observational data mainly collected from low-elevation floodplains and the over-flexibility of available models can negatively affect the verifiability of large-scale simulations of WTD. Ultimately, we thoroughly discuss future directions that may help hydrogeologists decide how to improve machine learning-based WTD estimations. In particular, we advocate for the use of proxy satellite data, the incorporation of physical laws, the implementation of better model verification standards, the development of novel globally-available emergent indices, and the collection of more reliable observations.</div></div>","PeriodicalId":7614,"journal":{"name":"Advances in Water Resources","volume":"201 ","pages":"Article 104955"},"PeriodicalIF":4.0,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143870566","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Image-well solution for island aquifers with pumping, recharge, and complex coastlines","authors":"Ying-Fan Lin ,&nbsp;Barret L. Kurylyk ,&nbsp;Adrian D. Werner ,&nbsp;Chih-Yu Liu ,&nbsp;Cristina Solórzano-Rivas ,&nbsp;Jun-Hong Lin","doi":"10.1016/j.advwatres.2025.104977","DOIUrl":"10.1016/j.advwatres.2025.104977","url":null,"abstract":"<div><div>This study presents an innovative mathematical framework that integrates a new analytical solution with the image-well method to model island aquifers under the combined influences of pumping, recharge, and complex coastline geometries. Past analytical solutions often rely on simplified boundary conditions and assume axially or radially symmetric coastline geometries, limiting their ability to address multiple stressors and complex island geometries. In contrast, the proposed framework leverages the computational speed and simplicity of analytical modeling while incorporating an image-well approach to accommodate boundaries with arbitrary shapes. To validate the method accuracy and robustness, convergence analyses and comparisons with an established analytical solution are conducted. Additionally, new indices are introduced to evaluate the sensitivity of the freshwater–saltwater interface depth to various forcings and to assess uncertainties in vulnerability indices. An illustrative case study, based loosely on Kinmen Island, Taiwan, is used to demonstrate the applicability of the approach to optimize the pumping rates for multiple wells while ensuring that the interface depth remains within safe limits. Overall, the presented methodology provides a flexible and efficient tool for groundwater resource management in coastal regions, enabling assessment of saltwater intrusion risk and informing sustainable water-use strategies for coastal regions under dynamic environmental conditions.</div></div>","PeriodicalId":7614,"journal":{"name":"Advances in Water Resources","volume":"201 ","pages":"Article 104977"},"PeriodicalIF":4.0,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143855860","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Multi-GPUs based SWE algorithm and its application in the simulation of flood routing","authors":"Yang-Yang Zhang ,&nbsp;Wen-Jie Xu ,&nbsp;Fu-Qiang Tian ,&nbsp;Xiao-Hu Du","doi":"10.1016/j.advwatres.2025.104985","DOIUrl":"10.1016/j.advwatres.2025.104985","url":null,"abstract":"<div><div>Hydrodynamic simulation based on shallow water wave equations (SWE) is one of the most useful methods for flood routing analysis. However, its widespread application in large-scale basin flood disaster prevention is hindered by significant computational efficiency challenges. Addressing this issue, a hydrodynamic algorithm based on multi-GPUs is provided, and the program named as CoSim-SWE is developed in this study. In the algorithm, the computational domain is partitioned into multiple subdomains, and assigned to each single GPU device to improve the computing power and efficiency. The accuracy of the multi-GPUs algorithm is validated based on two benchmarks. To demonstrate its practical utility, the developed CoSim-SWE is used to reconstruct the process of “11·03″ flood routing induced by the dam breach of Baige barrier dam on the Jinsha River, Sichuan province, China. The numerical results are compared with the field investigations. Furthermore, computational efficiency of the CoSim-SWE is also analyzed, which shows the developed multi-GPUs algorithm significantly can be better used for simulation of the large-scale flood routing with high-efficiency.</div></div>","PeriodicalId":7614,"journal":{"name":"Advances in Water Resources","volume":"201 ","pages":"Article 104985"},"PeriodicalIF":4.0,"publicationDate":"2025-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143842997","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Pore scale modeling of wettability impact on CO2 capillary and dissolution trapping in natural porous media","authors":"Jinlei Wang ,&nbsp;Yongfei Yang ,&nbsp;Qi Zhang ,&nbsp;Qi Wang ,&nbsp;Huaisen Song ,&nbsp;Hai Sun ,&nbsp;Lei Zhang ,&nbsp;Junjie Zhong ,&nbsp;Kai Zhang ,&nbsp;Jun Yao","doi":"10.1016/j.advwatres.2025.104982","DOIUrl":"10.1016/j.advwatres.2025.104982","url":null,"abstract":"<div><div>Understanding CO₂ capillary trapping and dissolution trapping behaviors in deep saline aquifers is essential for improvement of sequestration efficiency. This study investigated the impact of rock wettability on CO₂ capillary and dissolution trapping in a Ketton carbonate rock through direct numerical simulation. Based on experimentally measured CO₂-brine-rock physics data, we performed the pore-scale simulation of drainage process under four wettability conditions, followed by imbibition process under six sets of flow rate conditions. The dynamic evolution of CO₂ clusters and dissolved CO₂ distribution during drainage and imbibition processes was tracked simultaneously. Our results showed that wettability significantly influenced the morphology of trapped CO₂ clusters, consequently impacting the dissolution behavior. During drainage process, CO₂ saturation is higher in relatively hydrophobic rock but the dissolution capacity is lower because of the lower specific interfacial area. During imbibition process, the trapped scCO₂ saturation is also higher in relatively hydrophobic rock, and the number of CO₂ clusters is higher with more dispersed distribution. Increasing the flow rate can significantly reduce the residual CO₂ saturation and thus affect the dissolution efficiency. The function relationship between Sherwood number <em>Sh</em> calculated by mass transfer coefficient <span><math><mover><mrow><mi>k</mi></mrow><mo>‾</mo></mover></math></span> and flow rate-dependent Péclet number <em>Pe</em> under different wettability conditions was proposed. The CO₂ dissolution process is more affected by the flow rate under relative hydrophilic conditions.</div></div>","PeriodicalId":7614,"journal":{"name":"Advances in Water Resources","volume":"201 ","pages":"Article 104982"},"PeriodicalIF":4.0,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143842868","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Laboratory experiments of rotating stratified exchange flows over a sediment bed","authors":"M.R. Maggi ,&nbsp;E.J. Hopfinger ,&nbsp;J. Sommeria ,&nbsp;C. Adduce ,&nbsp;S. Viboud ,&nbsp;T. Valran ,&nbsp;M.E. Negretti","doi":"10.1016/j.advwatres.2025.104959","DOIUrl":"10.1016/j.advwatres.2025.104959","url":null,"abstract":"<div><div>We present a pioneering experimental study of stratified, rotating exchange flows interacting with a bottom, mobile sediment bed that simulates large estuaries. Two-dimensional velocity fields are coupled with bed scan that allows to reconstruct the bed morphology. The experiments span a large parameter range, notably laminar to turbulent Ekman layer regimes (<span><math><mrow><mn>33</mn><mo>&lt;</mo><mi>R</mi><msub><mrow><mi>e</mi></mrow><mrow><mi>δ</mi></mrow></msub><mo>&lt;</mo><mn>192</mn></mrow></math></span>) and Burger numbers (<span><math><mrow><mn>0</mn><mo>.</mo><mn>4</mn><mo>&lt;</mo><mi>B</mi><mi>u</mi><mo>&lt;</mo><mn>2</mn><mo>.</mo><mn>1</mn></mrow></math></span>). For low Burger numbers (<span><math><mrow><mi>B</mi><mi>u</mi><mo>&lt;</mo><mn>1</mn></mrow></math></span>), meandering of the lower salty layer occurs due to baroclinic instability, leading to the formation of columnar vortices, whose size and phase speed are in agreement with theory (Pedlosky, 2013). The Ekman layer thickness is well identified in the experiments over a smooth bed giving access to an eddy viscosity, hence the friction velocity <span><math><msub><mrow><mi>u</mi></mrow><mrow><mo>∗</mo></mrow></msub></math></span>, which is used as the characteristic velocity of sediment erosion. We show that sediment transport is driven mainly by Ekman dynamics with a net transport across the channel cross-section in direction of the geostrophic slope. For low Burger numbers, meandering induces further variability in cross- and along-channel velocities affecting sediment transport. The estimated non-dimensional bed-form wavelengths align closely with previous values of ripples in gravity currents reported in the literature. Notably, these wavelengths do not indicate a transition from ripples to dunes with increasing Yalin numbers, unlike in open channel flows. Finally, the sediment suspension model of Maggi et al. (2024) is extended by including across-channel sediment transport, that explains the observed temporal change in bed-forms and the effect of Ekman pumping on the suspended sediment layer thickness.</div></div>","PeriodicalId":7614,"journal":{"name":"Advances in Water Resources","volume":"200 ","pages":"Article 104959"},"PeriodicalIF":4.0,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143816019","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Likelihood-free inference and hierarchical data assimilation for geological carbon storage","authors":"Wenchao Teng,&nbsp;Louis J. Durlofsky","doi":"10.1016/j.advwatres.2025.104961","DOIUrl":"10.1016/j.advwatres.2025.104961","url":null,"abstract":"<div><div>Data assimilation will be essential for the management and expansion of geological carbon storage operations. In traditional data assimilation approaches a fixed set of geological hyperparameters, such as mean and standard deviation of log-permeability, is often assumed. Such hyperparameters, however, may be highly uncertain in practical CO<span><math><msub><mrow></mrow><mrow><mn>2</mn></mrow></msub></math></span> storage applications where measurements are scarce. In this study, we develop a hierarchical data assimilation framework for carbon storage that treats hyperparameters as uncertain variables characterized by hyperprior distributions. To deal with the computationally intractable likelihood function in hyperparameter estimation, we apply a likelihood-free (or simulation-based) inference algorithm, specifically sequential Monte Carlo-based approximate Bayesian computation (SMC-ABC), to draw posterior samples of hyperparameters given dynamic monitoring well data. In the second step we use an ensemble smoother with multiple data assimilation (ESMDA) procedure to provide posterior realizations of grid-block permeability. To reduce computational costs, a 3D recurrent R-U-Net deep learning-based surrogate model is applied for forward function evaluations. The accuracy of the surrogate model is established through comparisons to high-fidelity simulation results. A rejection sampling (RS) procedure for data assimilation is applied to provide reference posterior results. Detailed data assimilation results from SMC-ABC-ESMDA are compared to those from the reference RS method. These include marginal posterior distributions of hyperparameters, pairwise marginal posterior samples, and history matching results for pressure and saturation at the monitoring location. Close agreement is achieved with ‘converged’ RS results, for two synthetic true models, in all quantities considered. Importantly, the SMC-ABC-ESMDA procedure provides speedup of 1–2 orders of magnitude relative to RS for the two cases. A modified standalone ESMDA procedure, able to treat uncertain hyperparameters, is introduced for comparison purposes. For the same number of function evaluations, the hierarchical data assimilation approach is shown to provide superior results for posterior hyperparameter distributions and monitoring well pressure predictions.</div></div>","PeriodicalId":7614,"journal":{"name":"Advances in Water Resources","volume":"201 ","pages":"Article 104961"},"PeriodicalIF":4.0,"publicationDate":"2025-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143850612","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}