Journal of Hydrology最新文献_第6页

Exacerbating hydrological extremes in China’s large reservoir drainage areas

IF 5.9 1区地球科学 Q1 ENGINEERING, CIVIL

Journal of Hydrology

Pub Date : 2025-04-14 DOI: 10.1016/j.jhydrol.2025.133297

Xinyu Li , Kaiwen Wang , Changming Liu , Gang Zhao , Zhouyuqian Jiang , Qiuyu Luo , Guan Wang , Dan Zhang , Jiamiao Yu , Xiaomang Liu

Reservoirs are vital infrastructure for mitigating hydrological extremes, providing water during droughts, and reducing risks associated with floods. Under a warming climate, increasing hydrological extremes in upstream catchments threaten water supply sustainability and dam security. However, the evolution and drivers of these extremes are still poorly understood due to limited precise drainage boundary data. To address this gap, we combine a delineation algorithm with manual adjustments according to recorded drainage areas, creating the most comprehensive publicly available inventory of 907 large Chinese reservoirs, each with a storage capacity exceeding 0.1 km³. By integrating delineated boundaries with an observation-based China natural runoff dataset, we find nearly 40 % of reservoirs face more intense and frequent droughts, jeopardizing their role in supporting regional water transfer projects. Additionally, nearly 60 % experience worsening pluvial conditions, putting reservoirs in the northwest, northeast, and lower Yangtze regions under flood control and coordination pressures. These intensifying hydrological extremes strongly correlate with climate variability modes, while their variations are further influenced by climate change, widespread greening, and other external factors. Given reservoirs’ essential role in human water use, this study highlights the urgent need to understand the effects of climate and landscape changes to advance sustainable water resource management and safeguard water security.

{"title":"Exacerbating hydrological extremes in China’s large reservoir drainage areas","authors":"Xinyu Li , Kaiwen Wang , Changming Liu , Gang Zhao , Zhouyuqian Jiang , Qiuyu Luo , Guan Wang , Dan Zhang , Jiamiao Yu , Xiaomang Liu","doi":"10.1016/j.jhydrol.2025.133297","DOIUrl":"10.1016/j.jhydrol.2025.133297","url":null,"abstract":"<div><div>Reservoirs are vital infrastructure for mitigating hydrological extremes, providing water during droughts, and reducing risks associated with floods. Under a warming climate, increasing hydrological extremes in upstream catchments threaten water supply sustainability and dam security. However, the evolution and drivers of these extremes are still poorly understood due to limited precise drainage boundary data. To address this gap, we combine a delineation algorithm with manual adjustments according to recorded drainage areas, creating the most comprehensive publicly available inventory of 907 large Chinese reservoirs, each with a storage capacity exceeding 0.1 km<sup>3</sup>. By integrating delineated boundaries with an observation-based China natural runoff dataset, we find nearly 40 % of reservoirs face more intense and frequent droughts, jeopardizing their role in supporting regional water transfer projects. Additionally, nearly 60 % experience worsening pluvial conditions, putting reservoirs in the northwest, northeast, and lower Yangtze regions under flood control and coordination pressures. These intensifying hydrological extremes strongly correlate with climate variability modes, while their variations are further influenced by climate change, widespread greening, and other external factors. Given reservoirs’ essential role in human water use, this study highlights the urgent need to understand the effects of climate and landscape changes to advance sustainable water resource management and safeguard water security.</div></div>","PeriodicalId":362,"journal":{"name":"Journal of Hydrology","volume":"659 ","pages":"Article 133297"},"PeriodicalIF":5.9,"publicationDate":"2025-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143854419","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}

引用次数: 0

Estimating rainfall intensity from surveillance audio: A hybrid model-data-driven framework

IF 5.9 1区地球科学 Q1 ENGINEERING, CIVIL

Journal of Hydrology

Pub Date : 2025-04-13 DOI: 10.1016/j.jhydrol.2025.133295

Xing Wang , Kun Zhao , Haiqin Chen , Ang Zhou , Jiuwei Zhao , Shuaiyi Shi , Thomas Glade

Rainfall produces one of the most recognizable and variable sounds in nature. Audio data collected by widespread surveillance cameras provide a continuous record of rainfall events, which offers a potential opportunity for high spatiotemporal resolution rainfall estimation. However, surveillance audio (SA)¹ often contains complicated environmental noise that challenges the characterisation of rainfall and makes it difficult to obtain rainfall information from SA data. This study proposes a hybrid model-data-driven framework for the numerical estimation of rain intensity based on SA. The framework is implemented in two steps: 1) a convolutional neural network (CNN) and long short-term memory (LSTM) were used to learn the frequency and temporal characteristics of rain sound, respectively, and a novel parallel neural network (PNN) was constructed to determine rain categories (e.g., light, moderate, and heavy) or the categories of rain intensities, which enabled a coarse-grained rain intensity estimation. 2) Subsequently, the Root-Mean-Square Energy (RMS-Energy) of the audio clip was employed as the indicator, and a fine-grained rainfall intensity numerical calculation model based on SA data was built. Experimental results reveal that the PNN achieves optimal performance compared to some existing relevant models, indicating that the proposed PNN can effectively determine the rain category from urban SA data. Moreover, observation from real-world surveillance scenarios demonstrates that our method achieves an average relative error of 8.01%–25.68% in the cumulative rainfall estimation. This research sheds light on building a new low-cost and high-resolution rainfall observation network based on the existing surveillance camera recourses and providing valuable support to the current rainfall observation networks.

降雨是自然界中最易辨认、最多变的声音之一。广泛使用的监控摄像头收集的音频数据可连续记录降雨事件，为高时空分辨率降雨估算提供了潜在机会。然而，监控音频（SA）1 通常包含复杂的环境噪声，这对降雨的特征描述提出了挑战，也使得从监控音频数据中获取降雨信息变得困难。本研究提出了一种基于 SA 的雨强数值估算混合模型-数据驱动框架。该框架分两步实施：1) 使用卷积神经网络（CNN）和长短期记忆（LSTM）分别学习雨声的频率和时间特征，并构建一个新颖的并行神经网络（PNN）来确定雨量类别（如小雨、中雨和大雨）或雨强类别，从而实现粗粒度雨强估算。2) 随后，以音频片段的均方根能量（RMS-Energy）为指标，建立了基于 SA 数据的细粒度雨强数值计算模型。实验结果表明，与现有的一些相关模型相比，PNN 实现了最佳性能，表明所提出的 PNN 能够有效地从城市 SA 数据中确定雨量类别。此外，实际监控场景的观测结果表明，我们的方法在累积雨量估算方面实现了 8.01%-25.68% 的平均相对误差。这项研究为基于现有监控摄像机资源构建低成本、高分辨率的新型雨量观测网络提供了启示，也为当前的雨量观测网络提供了有价值的支持。

{"title":"Estimating rainfall intensity from surveillance audio: A hybrid model-data-driven framework","authors":"Xing Wang , Kun Zhao , Haiqin Chen , Ang Zhou , Jiuwei Zhao , Shuaiyi Shi , Thomas Glade","doi":"10.1016/j.jhydrol.2025.133295","DOIUrl":"10.1016/j.jhydrol.2025.133295","url":null,"abstract":"<div><div>Rainfall produces one of the most recognizable and variable sounds in nature. Audio data collected by widespread surveillance cameras provide a continuous record of rainfall events, which offers a potential opportunity for high spatiotemporal resolution rainfall estimation. However, surveillance audio (SA)<span><span><sup>1</sup></span></span> often contains complicated environmental noise that challenges the characterisation of rainfall and makes it difficult to obtain rainfall information from SA data. This study proposes a hybrid model-data-driven framework for the numerical estimation of rain intensity based on SA. The framework is implemented in two steps: 1) a convolutional neural network (CNN) and long short-term memory (LSTM) were used to learn the frequency and temporal characteristics of rain sound, respectively, and a novel parallel neural network (PNN) was constructed to determine rain categories (e.g., light, moderate, and heavy) or the categories of rain intensities, which enabled a coarse-grained rain intensity estimation. 2) Subsequently, the Root-Mean-Square Energy (RMS-Energy) of the audio clip was employed as the indicator, and a fine-grained rainfall intensity numerical calculation model based on SA data was built. Experimental results reveal that the PNN achieves optimal performance compared to some existing relevant models, indicating that the proposed PNN can effectively determine the rain category from urban SA data. Moreover, observation from real-world surveillance scenarios demonstrates that our method achieves an average relative error of 8.01%–25.68% in the cumulative rainfall estimation. This research sheds light on building a new low-cost and high-resolution rainfall observation network based on the existing surveillance camera recourses and providing valuable support to the current rainfall observation networks.</div></div>","PeriodicalId":362,"journal":{"name":"Journal of Hydrology","volume":"659 ","pages":"Article 133295"},"PeriodicalIF":5.9,"publicationDate":"2025-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143850423","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}

引用次数: 0

Surface-temperature silica springs of the eastern Great Artesian Basin – Hydrogeology and hydrochemistry

IF 5.9 1区地球科学 Q1 ENGINEERING, CIVIL

Journal of Hydrology

Pub Date : 2025-04-13 DOI: 10.1016/j.jhydrol.2025.133311

J.A. Webb , R.J. Fensham , B. Laffineur

An unusual north–south line of ∼ 50 springs along the eastern margin of the Eromanga Basin, northeastern Australia, discharge from the outcrop margin of the Hutton Sandstone aquifer, which is folded so the springs are fed by easterly groundwater flow, in contrast to the dominant westwards flow within the Eromanga Basin. This means that the effective recharge area of the Hutton Sandstone in this region is less than previously estimated. The springs occur as pools which represent water-table windows, with groundwater ‘streams’ flowing from one side to the other along subhorizontal joints within a near-surface silcrete layer developed on the Hutton Sandstone. The springs are recharged through fractures in the silcrete, feeding laminar groundwater flow through the Hutton Sandstone until its outcrop terminates. At this point flow transfers into the overlying silcrete as concentrated pathways probably localised along broad, shallow troughs in the silcrete beneath surface drainage lines. The springs are surrounded by white siliceous precipitates with a groundmass of intergrown amorphous silica and kaolinite; this may have been allophane originally. Most silica springs are geothermal, yet the eastern Alice Tableland springs have surface temperatures. The elevated dissolved Si levels in these springs are due to dissolution of relatively soluble silica microcrystallites within the silcrete through which the spring water flows. The lack of calcite precipitation from the springs reflects the low Ca concentrations of the groundwater, probably due to strong Ca uptake by plants.

{"title":"Surface-temperature silica springs of the eastern Great Artesian Basin – Hydrogeology and hydrochemistry","authors":"J.A. Webb , R.J. Fensham , B. Laffineur","doi":"10.1016/j.jhydrol.2025.133311","DOIUrl":"10.1016/j.jhydrol.2025.133311","url":null,"abstract":"<div><div>An unusual north–south line of ∼ 50 springs along the eastern margin of the Eromanga Basin, northeastern Australia, discharge from the outcrop margin of the Hutton Sandstone aquifer, which is folded so the springs are fed by easterly groundwater flow, in contrast to the dominant westwards flow within the Eromanga Basin. This means that the effective recharge area of the Hutton Sandstone in this region is less than previously estimated. The springs occur as pools which represent water-table windows, with groundwater ‘streams’ flowing from one side to the other along subhorizontal joints within a near-surface silcrete layer developed on the Hutton Sandstone. The springs are recharged through fractures in the silcrete, feeding laminar groundwater flow through the Hutton Sandstone until its outcrop terminates. At this point flow transfers into the overlying silcrete as concentrated pathways probably localised along broad, shallow troughs in the silcrete beneath surface drainage lines. The springs are surrounded by white siliceous precipitates with a groundmass of intergrown amorphous silica and kaolinite; this may have been allophane originally. Most silica springs are geothermal, yet the eastern Alice Tableland springs have surface temperatures. The elevated dissolved Si levels in these springs are due to dissolution of relatively soluble silica microcrystallites within the silcrete through which the spring water flows. The lack of calcite precipitation from the springs reflects the low Ca concentrations of the groundwater, probably due to strong Ca uptake by plants.</div></div>","PeriodicalId":362,"journal":{"name":"Journal of Hydrology","volume":"659 ","pages":"Article 133311"},"PeriodicalIF":5.9,"publicationDate":"2025-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143837847","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}

引用次数: 0

A review of machine learning, remote sensing, and statistical methods for reservoir water quality assessment

IF 5.9 1区地球科学 Q1 ENGINEERING, CIVIL

Journal of Hydrology

Pub Date : 2025-04-13 DOI: 10.1016/j.jhydrol.2025.133323

Mohammad Reza Nikoo , Abrar Al Aamri , Talal Etri , Ghazi Al-Rawas

Water reservoirs perform a number of essential functions, including water supply, flood control, hydropower generation, and agricultural and industrial support. In order to meet specific standards, the reservoir water quality needs to be protected. Because of human activities, including industrial discharges and agricultural runoff, reservoir’s water quality deteriorates. Deforestation and erosion in the upstream region exacerbate the problem, disrupting the ecology. A comprehensive management practice is necessary to maintain reservoir water quality in addition to changes in flow patterns, temperature changes, and nutrient enrichment. A number of methods have been employed, including Remote Sensing (RS) for spatial monitoring of environmental change, Machine Learning (ML) for estimation/predicting water quality, and Multivariate Statistical Analysis (MSA) that can identify relationships among water quality variables and patterns. By examining the strengths of these methods, it is possible to maximize the effectiveness of reservoir management. For instance, by understanding each method, it is possible to identify the optimal combination of techniques to achieve the best results. Furthermore, it addresses a wide range of challenges related to assessing water quality and ecosystem health. The use of one or more of these approaches will depend on the objectives, data characteristics, and resources available. Additionally, it can be used to identify and mitigate the risks associated with reservoir management. The articles in this review paper were limited to those published between 2000 and 2023, with a reasonable geographical distribution based on our literature search in the SCOPUS database.

{"title":"A review of machine learning, remote sensing, and statistical methods for reservoir water quality assessment","authors":"Mohammad Reza Nikoo , Abrar Al Aamri , Talal Etri , Ghazi Al-Rawas","doi":"10.1016/j.jhydrol.2025.133323","DOIUrl":"10.1016/j.jhydrol.2025.133323","url":null,"abstract":"<div><div>Water reservoirs perform a number of essential functions, including water supply, flood control, hydropower generation, and agricultural and industrial support. In order to meet specific standards, the reservoir water quality needs to be protected. Because of human activities, including industrial discharges and agricultural runoff, reservoir’s water quality deteriorates. Deforestation and erosion in the upstream region exacerbate the problem, disrupting the ecology. A comprehensive management practice is necessary to maintain reservoir water quality in addition to changes in flow patterns, temperature changes, and nutrient enrichment. A number of methods have been employed, including Remote Sensing (RS) for spatial monitoring of environmental change, Machine Learning (ML) for estimation/predicting water quality, and Multivariate Statistical Analysis (MSA) that can identify relationships among water quality variables and patterns. By examining the strengths of these methods, it is possible to maximize the effectiveness of reservoir management. For instance, by understanding each method, it is possible to identify the optimal combination of techniques to achieve the best results. Furthermore, it addresses a wide range of challenges related to assessing water quality and ecosystem health. The use of one or more of these approaches will depend on the objectives, data characteristics, and resources available. Additionally, it can be used to identify and mitigate the risks associated with reservoir management. The articles in this review paper were limited to those published between 2000 and 2023, with a reasonable geographical distribution based on our literature search in the SCOPUS database.</div></div>","PeriodicalId":362,"journal":{"name":"Journal of Hydrology","volume":"659 ","pages":"Article 133323"},"PeriodicalIF":5.9,"publicationDate":"2025-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143834618","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}

引用次数: 0

A crop-specific dynamic irrigation scheme in a regional land surface-hydrologic modeling framework for improving human water-use estimation and irrigation impact assessment

IF 5.9 1区地球科学 Q1 ENGINEERING, CIVIL

Journal of Hydrology

Pub Date : 2025-04-13 DOI: 10.1016/j.jhydrol.2025.133322

Qianya Yang , Jianhui Wei , Chuanguo Yang , Huanghe Gu , Jianyong Ma , Ningpeng Dong , Joël Arnault , Patrick Laux , Benjamin Fersch , Shasha Shang , Zhongbo Yu , Harald Kunstmann

Irrigation has a notable impact on the natural environment by changing the water and energy balance at the land surface and thereby altering atmospheric processes. Assessing these impacts and estimating irrigation water demand often involves using process-based models that incorporate the representation of irrigation practices. However, current irrigation schemes are primarily tailored to arid and semi-arid regions, and there is a research gap for humid multi-cropping rice regions. In response, this study introduces a Crop-specific Dynamic Irrigation (CDI) scheme, seamlessly integrated into the land surface-hydrologic model NOAH-HMS. This development enables the differentiation of irrigation practices for rice and non-rice crops, facilitating more accurate estimates of water demand for irrigation. The newly developed model is applied to an important cropping region in southern China, the Poyang Lake Basin (PLB), where the rice cultivation area accounts for over 60% of all crop cultivation. Compared to the widely used traditional Dynamic Irrigation (DI) scheme, integrating CDI into NOAH-HMS improves the model performance in simulating irrigation water amount over the PLB, with a mean relative error between 2007–2015 reduced by 39%, and a correlation coefficient increased by +0.26. The identified impacts on the surface water and energy balance are more pronounced at local scale, especially over the intensively irrigated areas. The performed interannual variability analysis demonstrates that our irrigation scheme CDI developed in this study allows to estimate irrigation water use under different drought conditions and has the applicability of mitigating risks of crop failures due to for example compound dry and hot. We conclude that our Crop-specific Dynamic Irrigation scheme is highly advantageous for multi-cropping rice regions and holds the potential for expansion into the fully coupled atmospheric-hydrologic systems with a more comprehensive representation of human activities.

{"title":"A crop-specific dynamic irrigation scheme in a regional land surface-hydrologic modeling framework for improving human water-use estimation and irrigation impact assessment","authors":"Qianya Yang , Jianhui Wei , Chuanguo Yang , Huanghe Gu , Jianyong Ma , Ningpeng Dong , Joël Arnault , Patrick Laux , Benjamin Fersch , Shasha Shang , Zhongbo Yu , Harald Kunstmann","doi":"10.1016/j.jhydrol.2025.133322","DOIUrl":"10.1016/j.jhydrol.2025.133322","url":null,"abstract":"<div><div>Irrigation has a notable impact on the natural environment by changing the water and energy balance at the land surface and thereby altering atmospheric processes. Assessing these impacts and estimating irrigation water demand often involves using process-based models that incorporate the representation of irrigation practices. However, current irrigation schemes are primarily tailored to arid and semi-arid regions, and there is a research gap for humid multi-cropping rice regions. In response, this study introduces a Crop-specific Dynamic Irrigation (CDI) scheme, seamlessly integrated into the land surface-hydrologic model NOAH-HMS. This development enables the differentiation of irrigation practices for rice and non-rice crops, facilitating more accurate estimates of water demand for irrigation. The newly developed model is applied to an important cropping region in southern China, the Poyang Lake Basin (PLB), where the rice cultivation area accounts for over 60% of all crop cultivation. Compared to the widely used traditional Dynamic Irrigation (DI) scheme, integrating CDI into NOAH-HMS improves the model performance in simulating irrigation water amount over the PLB, with a mean relative error between 2007–2015 reduced by 39%, and a correlation coefficient increased by +0.26. The identified impacts on the surface water and energy balance are more pronounced at local scale, especially over the intensively irrigated areas. The performed interannual variability analysis demonstrates that our irrigation scheme CDI developed in this study allows to estimate irrigation water use under different drought conditions and has the applicability of mitigating risks of crop failures due to for example compound dry and hot. We conclude that our Crop-specific Dynamic Irrigation scheme is highly advantageous for multi-cropping rice regions and holds the potential for expansion into the fully coupled atmospheric-hydrologic systems with a more comprehensive representation of human activities.</div></div>","PeriodicalId":362,"journal":{"name":"Journal of Hydrology","volume":"659 ","pages":"Article 133322"},"PeriodicalIF":5.9,"publicationDate":"2025-04-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143851740","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}

引用次数: 0

Multi-scale dynamic spatiotemporal graph attention network for forecasting karst spring discharge

IF 5.9 1区地球科学 Q1 ENGINEERING, CIVIL

Journal of Hydrology

Pub Date : 2025-04-12 DOI: 10.1016/j.jhydrol.2025.133289

Renjie Zhou

Karst aquifers are important groundwater resources that supply drinking water for approximately 25 % of the world’s population. Their complex hydrogeological structures, dual-flow regimes, and highly heterogeneous flow pose significant challenges for accurate hydrodynamic modeling and sustainable management. Traditional modeling approaches often struggle to capture the intricate spatial dependencies and multi-scale temporal patterns inherent in karst systems, particularly the interactions between rapid conduit flow and slower matrix flow. This study proposes a novel multi-scale dynamic graph attention network integrated with long short-term memory model (GAT-LSTM) to innovatively learn and integrate spatial and temporal dependencies in karst systems for forecasting spring discharge. The model introduces several innovative components: (1) graph-based neural networks with dynamic edge-weighting mechanism are proposed to learn and update spatial dependencies based on both geographic distances and learned hydrological relationships, (2) a multi-head attention mechanism is adopted to capture different aspects of spatial relationships simultaneously, and (3) a hierarchical temporal architecture is incorporated to process hydrological temporal patterns at both monthly and seasonal scales with an adaptive fusion mechanism for final results. These features enable the proposed model to effectively account for the dual-flow dynamics in karst systems, where rapid conduit flow and slower matrix flow coexist. The newly proposed model is applied to the Barton Springs of the Edwards Aquifer in Texas. The results demonstrate that it can obtain more accurate and robust prediction performance across various time steps compared to traditional temporal and spatial deep learning approaches. Based on the multi-scale GAT-LSTM model, a comprehensive ablation analysis and permutation feature important are conducted to analyze the relative contribution of various input variables on the final prediction. These findings highlight the intricate nature of karst systems and demonstrate that effective spring discharge prediction requires comprehensive monitoring networks encompassing both primary recharge contributors and supplementary hydrological features that may serve as valuable indicators of system-wide conditions.

{"title":"Multi-scale dynamic spatiotemporal graph attention network for forecasting karst spring discharge","authors":"Renjie Zhou","doi":"10.1016/j.jhydrol.2025.133289","DOIUrl":"10.1016/j.jhydrol.2025.133289","url":null,"abstract":"<div><div>Karst aquifers are important groundwater resources that supply drinking water for approximately 25 % of the world’s population. Their complex hydrogeological structures, dual-flow regimes, and highly heterogeneous flow pose significant challenges for accurate hydrodynamic modeling and sustainable management. Traditional modeling approaches often struggle to capture the intricate spatial dependencies and multi-scale temporal patterns inherent in karst systems, particularly the interactions between rapid conduit flow and slower matrix flow. This study proposes a novel multi-scale dynamic graph attention network integrated with long short-term memory model (GAT-LSTM) to innovatively learn and integrate spatial and temporal dependencies in karst systems for forecasting spring discharge. The model introduces several innovative components: (1) graph-based neural networks with dynamic edge-weighting mechanism are proposed to learn and update spatial dependencies based on both geographic distances and learned hydrological relationships, (2) a multi-head attention mechanism is adopted to capture different aspects of spatial relationships simultaneously, and (3) a hierarchical temporal architecture is incorporated to process hydrological temporal patterns at both monthly and seasonal scales with an adaptive fusion mechanism for final results. These features enable the proposed model to effectively account for the dual-flow dynamics in karst systems, where rapid conduit flow and slower matrix flow coexist. The newly proposed model is applied to the Barton Springs of the Edwards Aquifer in Texas. The results demonstrate that it can obtain more accurate and robust prediction performance across various time steps compared to traditional temporal and spatial deep learning approaches. Based on the multi-scale GAT-LSTM model, a comprehensive ablation analysis and permutation feature important are conducted to analyze the relative contribution of various input variables on the final prediction. These findings highlight the intricate nature of karst systems and demonstrate that effective spring discharge prediction requires comprehensive monitoring networks encompassing both primary recharge contributors and supplementary hydrological features that may serve as valuable indicators of system-wide conditions.</div></div>","PeriodicalId":362,"journal":{"name":"Journal of Hydrology","volume":"659 ","pages":"Article 133289"},"PeriodicalIF":5.9,"publicationDate":"2025-04-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143837846","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}

引用次数: 0

Enhancement of the SWAT+ model for simulating paddy rice cultivation and irrigation management in agricultural watersheds

IF 5.9 1区地球科学 Q1 ENGINEERING, CIVIL

Journal of Hydrology

Pub Date : 2025-04-12 DOI: 10.1016/j.jhydrol.2025.133288

Jaehak Jeong , Jeffery Arnold , Seonggyu Park , Ricardo Sorando , Soon-Kun Choi , Min-Kyeong Kim

Paddy cultivation accounts for over two-thirds of global rice production and 21 % of agricultural irrigation. While SWAT+ shows potential for simulating hydrology in paddy-dominant watersheds, improvements are needed. This study enhances SWAT+ by introducing a process module to simulate paddy hydrology and irrigation management. Unlike conventional hydrologic settings for Hydrologic Response Units (HRUs), the paddy module developed in SWAT+ introduces hydrologic mass balance for standing water in paddy HRUs. To implement paddy management, including transplanting, puddling, paddy irrigation, and fertilizer application, new paddy-specific conditions and actions are incorporated into SWAT+ decision tables. Case studies in South Korea and Spain demonstrate significant improvements in streamflow prediction and irrigation volume estimation. The Potential EvapoTranspiration COefficient (PETCO) was the most sensitive parameter in both watersheds, while the PERcolation Coefficient (PERCO) was more influential in non-paddy areas with high percolation rates. The study highlights distinct water balance differences, with paddy fields exhibiting higher evapotranspiration (>75 %) and surface runoff (>175 %) than other land uses. Compared to the curve number method, the paddy module improved streamflow simulation, achieving NSE values of 0.7–0.84 and PBIAS within ±10 %, particularly capturing high flows during the growing season. These enhancements strengthen SWAT+’s applicability for paddy-dominant watersheds, offering valuable insights for agricultural hydrology research.

{"title":"Enhancement of the SWAT+ model for simulating paddy rice cultivation and irrigation management in agricultural watersheds","authors":"Jaehak Jeong , Jeffery Arnold , Seonggyu Park , Ricardo Sorando , Soon-Kun Choi , Min-Kyeong Kim","doi":"10.1016/j.jhydrol.2025.133288","DOIUrl":"10.1016/j.jhydrol.2025.133288","url":null,"abstract":"<div><div>Paddy cultivation accounts for over two-thirds of global rice production and 21 % of agricultural irrigation. While SWAT+ shows potential for simulating hydrology in paddy-dominant watersheds, improvements are needed. This study enhances SWAT+ by introducing a process module to simulate paddy hydrology and irrigation management. Unlike conventional hydrologic settings for Hydrologic Response Units (HRUs), the paddy module developed in SWAT+ introduces hydrologic mass balance for standing water in paddy HRUs. To implement paddy management, including transplanting, puddling, paddy irrigation, and fertilizer application, new paddy-specific conditions and actions are incorporated into SWAT+ decision tables. Case studies in South Korea and Spain demonstrate significant improvements in streamflow prediction and irrigation volume estimation. The Potential EvapoTranspiration COefficient (PETCO) was the most sensitive parameter in both watersheds, while the PERcolation Coefficient (PERCO) was more influential in non-paddy areas with high percolation rates. The study highlights distinct water balance differences, with paddy fields exhibiting higher evapotranspiration (>75 %) and surface runoff (>175 %) than other land uses. Compared to the curve number method, the paddy module improved streamflow simulation, achieving NSE values of 0.7–0.84 and PBIAS within ±10 %, particularly capturing high flows during the growing season. These enhancements strengthen SWAT+’s applicability for paddy-dominant watersheds, offering valuable insights for agricultural hydrology research.</div></div>","PeriodicalId":362,"journal":{"name":"Journal of Hydrology","volume":"659 ","pages":"Article 133288"},"PeriodicalIF":5.9,"publicationDate":"2025-04-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143844926","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}

引用次数: 0

Floodplain hydrodynamics and connectivity in a natural compound channel during unsteady flow events

IF 5.9 1区地球科学 Q1 ENGINEERING, CIVIL

Journal of Hydrology

Pub Date : 2025-04-12 DOI: 10.1016/j.jhydrol.2025.133305

Jiaming Liu , Yang Xiao , Saiyu Yuan , Taotao Zhang , Qingwei Lin , Kang Yuan , Ruiqi Wang , Carlo Gualtieri

This study investigates the hydrodynamics features and the floodplain connectivity of a natural compound channel under unsteady flood conditions using a two-dimensional shallow water equation model and a Lagrangian particle tracking method. Two flood events in 2019 in the middle reach of the Ganjiang River in China were simulated. The results show that during the rising stage, flow first passed through the low-lying areas of the floodplain. The floodplain discharge ratio increased almost linearly with the depth ratio between the floodplain and main channel when the floodplain was not fully inundated. When the floodplain was fully inundated, a second linear relationship was found between the floodplain discharge ratio and depth ratio. During the falling stage, flow first moved back to the low-lying floodplain and main channel before fully receding from the floodplain. The sequence of peak velocity, discharge and stage in unsteady flow lee to higher velocities, lower depths, and shorter residence times during the rising limb compared to those in the falling limb at the same discharge. The threshold discharge for floodplain inundation was during the rising stage larger than during the falling stage. The shortest particle residence time was observed at the flood peak, while the residence time in the rising stage was longer than in the falling stage. The particle travel distance was similar at different stages. The exchange flux between the river and floodplain increased with inflow discharge following a power law relationship. The ratio of exchange flux to inflow discharge also increased with inflow discharge up to an upper limit of 65.5 %. Particle residence time was negatively correlated with discharge following a power law with a lower limit of 2630 s, while particle travel distance is positively correlated with discharge following a power law with an upper limit of 2325 m. These findings shed light on the complex hydrodynamic processes and connectivity patterns in natural compound channels during unsteady flood conditions.

{"title":"Floodplain hydrodynamics and connectivity in a natural compound channel during unsteady flow events","authors":"Jiaming Liu , Yang Xiao , Saiyu Yuan , Taotao Zhang , Qingwei Lin , Kang Yuan , Ruiqi Wang , Carlo Gualtieri","doi":"10.1016/j.jhydrol.2025.133305","DOIUrl":"10.1016/j.jhydrol.2025.133305","url":null,"abstract":"<div><div>This study investigates the hydrodynamics features and the floodplain connectivity of a natural compound channel under unsteady flood conditions using a two-dimensional shallow water equation model and a Lagrangian particle tracking method. Two flood events in 2019 in the middle reach of the Ganjiang River in China were simulated. The results show that during the rising stage, flow first passed through the low-lying areas of the floodplain. The floodplain discharge ratio increased almost linearly with the depth ratio between the floodplain and main channel when the floodplain was not fully inundated. When the floodplain was fully inundated, a second linear relationship was found between the floodplain discharge ratio and depth ratio. During the falling stage, flow first moved back to the low-lying floodplain and main channel before fully receding from the floodplain. The sequence of peak velocity, discharge and stage in unsteady flow lee to higher velocities, lower depths, and shorter residence times during the rising limb compared to those in the falling limb at the same discharge. The threshold discharge for floodplain inundation was during the rising stage larger than during the falling stage. The shortest particle residence time was observed at the flood peak, while the residence time in the rising stage was longer than in the falling stage. The particle travel distance was similar at different stages. The exchange flux between the river and floodplain increased with inflow discharge following a power law relationship. The ratio of exchange flux to inflow discharge also increased with inflow discharge up to an upper limit of 65.5 %. Particle residence time was negatively correlated with discharge following a power law with a lower limit of 2630 s, while particle travel distance is positively correlated with discharge following a power law with an upper limit of 2325 m. These findings shed light on the complex hydrodynamic processes and connectivity patterns in natural compound channels during unsteady flood conditions.</div></div>","PeriodicalId":362,"journal":{"name":"Journal of Hydrology","volume":"659 ","pages":"Article 133305"},"PeriodicalIF":5.9,"publicationDate":"2025-04-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143848547","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}

引用次数: 0

Multi-Phase mobilization and degradation pathway of natural organic matter within alluvial-lacustrine aquifer sediments

IF 5.9 1区地球科学 Q1 ENGINEERING, CIVIL

Journal of Hydrology

Pub Date : 2025-04-12 DOI: 10.1016/j.jhydrol.2025.133310

Wenkai Qiu , Yao Du , Wenhui Liu , Xinyu Liu , Yamin Deng , Teng Ma , Yanxin Wang

Alluvial-lacustrine sediments are crucial organic matter pools in the subsurface, fueling biogeochemical processes in aquifers through the release of dissolved organic matter (DOM). However, little is known about the detailed mobilization and degradation pathways of organic matter from sediments to groundwater. Here, we examined porewater bridging sediments and groundwater and characterized the fluorescence properties and molecular composition of DOM in porewater with varying binding strengths to sediments. Results showed that bound water DOM primarily contained polycyclic aromatics, polyphenols, and highly unsaturated compounds with high O/C ratios and humification indices. Free water DOM, by contrast, was rich in aliphatic compounds with high H/C ratios and biological indices. Unique molecular formulae analysis revealed that in bound water, CHO + nN polyphenolic compounds were consumed, and highly unsaturated CHO + 2 N compounds were produced; in free water, highly unsaturated CHO + 2 N compounds were consumed, and CHO + 1 N aliphatic compounds were produced. This molecular fractionation of organic matter is likely controlled by abiotic diffusion and adsorption as well as by biotic microbial activity, with the former dominating in bound water and the latter in free water. This may relate to the stable double-layer structure in bound water and the increased exposure to electron acceptors in free water. Together, these dual mechanisms produce a high proportion (>50 %) of biodegradable DOM in free water, potentially explaining the geogenic enrichment of ammonium in local aquifers. Our results highlight the strong influence of local mobilization and degradation processes on the molecular characteristics of DOM, bearing significant environmental implications for the evolution of groundwater quality.

冲积-湖积沉积物是地下重要的有机物库，通过释放溶解有机物（DOM）促进含水层的生物地球化学过程。然而，人们对有机物从沉积物到地下水的详细迁移和降解途径知之甚少。在此，我们研究了连接沉积物和地下水的孔隙水，并描述了与沉积物结合强度不同的孔隙水中 DOM 的荧光特性和分子组成。结果表明，结合水 DOM 主要含有多环芳烃、多酚和高度不饱和化合物，具有较高的 O/C 比和腐殖化指数。相比之下，自由水 DOM 则富含脂肪族化合物，具有较高的 H/C 比和生物指数。独特的分子式分析表明，在结合水中，消耗的是 CHO + nN 多酚化合物，产生的是高度不饱和的 CHO + 2 N 化合物；在自由水中，消耗的是高度不饱和的 CHO + 2 N 化合物，产生的是 CHO + 1 N 脂肪族化合物。有机物的这种分子分馏可能受到非生物扩散和吸附以及生物微生物活动的控制，前者在结合水中占主导地位，后者在自由水中占主导地位。这可能与结合水中稳定的双层结构和自由水中更多的电子受体有关。这两种机制共同作用，在自由水中产生了很高比例（50%）的可生物降解的 DOM，这可能解释了当地含水层中铵的地质富集。我们的研究结果凸显了当地迁移和降解过程对 DOM 分子特征的巨大影响，对地下水质量的演变具有重要的环境意义。

{"title":"Multi-Phase mobilization and degradation pathway of natural organic matter within alluvial-lacustrine aquifer sediments","authors":"Wenkai Qiu , Yao Du , Wenhui Liu , Xinyu Liu , Yamin Deng , Teng Ma , Yanxin Wang","doi":"10.1016/j.jhydrol.2025.133310","DOIUrl":"10.1016/j.jhydrol.2025.133310","url":null,"abstract":"<div><div>Alluvial-lacustrine sediments are crucial organic matter pools in the subsurface, fueling biogeochemical processes in aquifers through the release of dissolved organic matter (DOM). However, little is known about the detailed mobilization and degradation pathways of organic matter from sediments to groundwater. Here, we examined porewater bridging sediments and groundwater and characterized the fluorescence properties and molecular composition of DOM in porewater with varying binding strengths to sediments. Results showed that bound water DOM primarily contained polycyclic aromatics, polyphenols, and highly unsaturated compounds with high O/C ratios and humification indices. Free water DOM, by contrast, was rich in aliphatic compounds with high H/C ratios and biological indices. Unique molecular formulae analysis revealed that in bound water, CHO + nN polyphenolic compounds were consumed, and highly unsaturated CHO + 2 N compounds were produced; in free water, highly unsaturated CHO + 2 N compounds were consumed, and CHO + 1 N aliphatic compounds were produced. This molecular fractionation of organic matter is likely controlled by abiotic diffusion and adsorption as well as by biotic microbial activity, with the former dominating in bound water and the latter in free water. This may relate to the stable double-layer structure in bound water and the increased exposure to electron acceptors in free water. Together, these dual mechanisms produce a high proportion (>50 %) of biodegradable DOM in free water, potentially explaining the geogenic enrichment of ammonium in local aquifers. Our results highlight the strong influence of local mobilization and degradation processes on the molecular characteristics of DOM, bearing significant environmental implications for the evolution of groundwater quality.</div></div>","PeriodicalId":362,"journal":{"name":"Journal of Hydrology","volume":"659 ","pages":"Article 133310"},"PeriodicalIF":5.9,"publicationDate":"2025-04-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143829726","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}

引用次数: 0

New insights into the solute transport processes of red soil combining CT scanning and stable isotope tracing

IF 5.9 1区地球科学 Q1 ENGINEERING, CIVIL

Journal of Hydrology

Pub Date : 2025-04-11 DOI: 10.1016/j.jhydrol.2025.133286

Xinni Ju , Lei Gao , Dongli She , Yuhua Jia , Asim Biswas , Xinhua Peng

Understanding soil pore structure’s influence on material transport is crucial for nutrient cycling and pollutant management, particularly in complex hydrological environments. This study investigated the relationship between soil structure and solute transport across three land uses (forest, citrus, and rainfed cropland) in China’s red soil region using X-ray computed tomography and water tracers. Forest soils exhibited well-connected macropores, while rainfed cropland showed numerous discrete micropores. Citrus land displayed heterogeneous pore distribution with high surface macroporosity but low overall macropore content. Breakthrough curves revealed three distinct transport stages: initial transport, rapid replacement, and complete stages, each demonstrating unique characteristics based on land use. The initial transport varied significantly among land uses (0.2 pore volumes for forest/citrus, 0.1 for cropland), with cropland showing nearly double the porosity (24.4 %) compared to other land uses at this stage. Complete stage convergence occurred at approximately 1.6 pore volumes across all land uses. The Continuous Time Random Walk model outperformed the Advection-Dispersion Equation in predicting solute transport (R² = 0.999), particularly in capturing late-time tailing phenomena. Bulk density, mean diameter, soil organic matter, and pore connectivity explained 51.6 % and 61.5 % of the variability in transport parameters ΦA and ΦB, respectively. These findings enhance our understanding of structure-transport relationships in red soil regions and provide valuable insights for sustainable land management practices, pollutant transport prediction, and the development of targeted soil remediation strategies in similar hydrological conditions worldwide.

{"title":"New insights into the solute transport processes of red soil combining CT scanning and stable isotope tracing","authors":"Xinni Ju , Lei Gao , Dongli She , Yuhua Jia , Asim Biswas , Xinhua Peng","doi":"10.1016/j.jhydrol.2025.133286","DOIUrl":"10.1016/j.jhydrol.2025.133286","url":null,"abstract":"<div><div>Understanding soil pore structure’s influence on material transport is crucial for nutrient cycling and pollutant management, particularly in complex hydrological environments. This study investigated the relationship between soil structure and solute transport across three land uses (forest, citrus, and rainfed cropland) in China’s red soil region using X-ray computed tomography and water tracers. Forest soils exhibited well-connected macropores, while rainfed cropland showed numerous discrete micropores. Citrus land displayed heterogeneous pore distribution with high surface macroporosity but low overall macropore content. Breakthrough curves revealed three distinct transport stages: initial transport, rapid replacement, and complete stages, each demonstrating unique characteristics based on land use. The initial transport varied significantly among land uses (0.2 pore volumes for forest/citrus, 0.1 for cropland), with cropland showing nearly double the porosity (24.4 %) compared to other land uses at this stage. Complete stage convergence occurred at approximately 1.6 pore volumes across all land uses. The Continuous Time Random Walk model outperformed the Advection-Dispersion Equation in predicting solute transport (R<sup>2</sup> = 0.999), particularly in capturing late-time tailing phenomena. Bulk density, mean diameter, soil organic matter, and pore connectivity explained 51.6 % and 61.5 % of the variability in transport parameters ΦA and ΦB, respectively. These findings enhance our understanding of structure-transport relationships in red soil regions and provide valuable insights for sustainable land management practices, pollutant transport prediction, and the development of targeted soil remediation strategies in similar hydrological conditions worldwide.</div></div>","PeriodicalId":362,"journal":{"name":"Journal of Hydrology","volume":"659 ","pages":"Article 133286"},"PeriodicalIF":5.9,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143834617","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}

引用次数: 0