Search

Hydraulic connections between a river and an adjacent aquifer are controlled by the river resistance and aquifer diffusivity. In this paper, we derive a spectral solution linking the power spectrum of river stage fluctuations to that of the hydraulic head of a confined aquifer by means of a physical scaling factor. The physical scaling factor represents an algebraic expression of the river resistance and aquifer diffusivity and is included in an exact spectral solution derived herein. Statistical measures of the aquifer diffusivity and river resistance are provided by fitting the solution versus observed groundwater hydraulic head obtained at several distances and/or frequencies. At a study site in the middle reach of the Yangtze River and downstream of the Three Gorges Dam in China, we find systematic damping of the hydraulic head variations with distance from the river, which follows a fractal pattern driven by the river stage. In general, the estimated parameters are consistent with results reported in the literature, which supports the validity of the proposed spectral approach, although the paper discusses advantages and limitations due to application conditions. Plain Language Summary: The characterization of hydraulic connections between a river and an adjacent aquifer requires knowledge of two critical parameters: the river resistance and the aquifer diffusivity. In this paper, we derive a spectral solution to link the power spectrum of river stage fluctuations to that of the groundwater hydraulic head in a confined aquifer and show how such solutions can be used to simultaneously assess the two critical parameters. Parameters estimated from the field case located in the middle reach of the Yangtze River and downstream of the Three Gorges Dam in China are consistent with results reported in the literature. Application conditions, advantages, and limitations of this spectral approach are also discussed. Key Points: The power spectrum of the hydraulic head is linked to that of the river stage by a scaling factorRiver resistance and aquifer diffusivity are analyzed in spectral formsTheoretical developments are applied to a field case study of the Yangtze River, China [ABSTRACT FROM AUTHOR]

Reliable urban flood modeling is highly demanded in emergency response, risk management, and urban planning related to urban flooding. In this paper, the Storm Water Management Model (SWMM) is adapted to simulate urban rainfall‐runoff and pipe drainage processes within the Dominant river tracing‐Routing Integrated with VIC Environment (DRIVE) model which accounts for natural river basin runoff generation and routing processes. The integrated DRIVE‐SWMM model (referred to as DRIVE‐Urban) allows to explicitly delineate the mass‐energy interactions between urban drainage system (e.g., pipes and dikes) and river networks. This presents a further step model development for accurate urban flooding prediction which is lacking in existing urban flood models and traditional hydrological models. The validity of the DRIVE‐Urban model is evaluated for three case studies in Haikou City, China, with camera observations of street inundation during typhoon landfalls and heavy rainfall events. The results show that the DRIVE‐Urban model successfully captures 62%, 69%, and 77% of the total observed inundated road‐sections for the three cases respectively. The third case study with severe flooding situation shows that the DRIVE‐Urban performance is further improved when given reliable river and tidal level information, indicating the importance of integrating river‐basin with urban hydrological and hydraulic modeling. Plain Language Summary: This study mainly focuses on the development of an urban flood model (Dominant river tracing‐Routing Integrated with VIC Environment, DRIVE‐Urban), which couples the river‐basin and urban hydrological modeling, with a great potential for large‐scale real‐time urban flood monitoring and forecasting through the current Global Flood Monitoring System. Two inundation algorithms are developed within the DRIVE‐Urban model. One is for the river inundation caused urban flooding, and the other one is for the urban inundation caused by manhole overflow or excessive runoff drainage over the city land surface. Multi‐source observations including the closed‐circuit television video, citizen photos and municipal reports are collected for model validation. The results show that the DRIVE‐Urban model simulates well the different type of urban inundations at the 5‐m high spatial resolution with a promising hit success (i.e., POD >60%) according to three case studies. In particular, the results for the typhoon Kalmaegi case show that the model performance can be improved significantly by coupling the upstream river flow simulation with the urban runoff‐routing process, which validates our strategy of including the upstream river‐basin hydrological process in the urban flood modeling. Key Points: The Dominant river tracing‐Routing Integrated with VIC Environment (DRIVE)‐Urban flood model is developed by coupling river basin and urban hydrological and hydraulic processTwo 2D inundation algorithms are implemented within the model to simulate the manhole inundating and river overflow flooding respectivelyThe validity of the DRIVE‐Urban model is well established through three case studies with observations during typhoon and rainfall events [ABSTRACT FROM AUTHOR]

Since the 2008 Wenchuan (Ms. 8.0) Earthquake, the foreland rivers of the Longmen Mountains have suffered from significant bed degradation, among which the Shi‐ting River has experienced the largest local degradation of more than 20 m in 7 years. Potential reasons of the dramatic degradation include: (a) sediment disconnectivity due to in‐channel weirs; (b) the mobilization effect on gravel of an increased sand supply as a result of earthquake‐induced landslides; and (c) sediment extraction due to intensive mining. In this paper, we study the complex interaction among the above‐mentioned factors in the Shi‐ting River, using a one‐dimensional river morphodynamic model. Simulation results show that in‐channel weirs can reduce bedload transport and lead to bed degradation that is proportional to weir height. When coupled with additional sand supply, the weirs preferentially trap gravel and deliver sand, augmenting the downstream mobility of gravel and thus the degradation. For the Shi‐ting River, the simulated bed degradation agrees well with the observation when an annual sediment mining of 16 million tons is implemented in the simulation, along with the effects of in‐channel weirs and sand supply. The contribution of sediment mining is one order of magnitude larger than the coupling effect of weirs and sand supply. Both the simulation and observation show that the largest bed degradation occurs downstream of the Renmin Weir, due to the large spatial interval between the Renmin Weir and the next grade control structure. Key Points: Sediment mining is likely the key factor causing dramatic bed degradation in the Shi‐ting River, ChinaThis degradation was abetted, but not primarily caused by earthquake‐derived sand mobilizing gravel, as proposed previouslyWeirs preferentially trap gravel and deliver sand, which enhances the mobility of gravel by sand [ABSTRACT FROM AUTHOR]

Agricultural production is accompanied by a large amount of water consumption, nonpoint source pollution, and greenhouse gas emissions. However, the comprehensive and quantitative analysis of associated impacts on regional water, the environment, and the economy caused by variations in agricultural distribution is insufficient. This paper evaluates the evolution of grain production distribution and its effects on water resources, the economy, and the environment in China by using virtual water theory. The results show that the grain production area located in northern China is characterized by scarce water resources and a less developed economy. Due to the imbalance between grain supply and demand, virtual water embedded in grain will transfer among regions. These flows have formed a pattern where virtual water transfers from the water‐scarce northern region to the water‐rich southern region, increasing from 72.99 Gm3 in 1997 to 124.64 Gm3 in 2014. Evolution of grain production distribution changes the spatial pattern of grain production and consumption, and it exacerbates water resource pressure, the gray water footprint, and greenhouse gas emissions in the area that exports grain virtual water. The gray water footprint and carbon emissions in the grain export area increased by 10.66% and 31.06% during the study period, respectively. Meanwhile, the distribution of regional grain production influences the allocation of water resources in agriculture and other industries. Due to the difference between the economic benefits created by industry and agriculture, grain virtual water flow will have effects on the regional economic development. Key Points: Grain virtual water flow (VWF) dynamics along with the variations in the grain planting distribution from long time series are exploredThe associated impacts on regional water, the environment, and the economy caused by grain virtual water flow are evaluatedA framework to evaluate economic impact of VWF is established by scenario analysis regarding industrial structure and water reallocation [ABSTRACT FROM AUTHOR]

Interbasin water transfers have been undertaken in those regions that water demand has gone beyond regional water availability. Some reservoirs in a recipient region have both natural inflow (NI) and the controlled inflow (CI) through water transfer. This study establishes a theoretical framework to analyze reservoir operation with combined NI and CI, via an optimization model to maximize the net benefit of reservoir operation, considering both deterministic and stochastic NI. Operation rules for both water transfer and reservoir release are derived from the framework. It is illustrated that the amount of CI depends on water availability to the recipient reservoir, the relationship between marginal benefit and marginal cost, the water transfer capacity, and the uncertainty level of NI. As forecast uncertainty increases, more water is transferred, and more water is left in the reservoir as carryover storage when the initial water availability (WA0) is small; however, with a certain high level of WA0 and forecast uncertainty, water transfer will be controlled to reduce the risk of water abandonment. Notably, the critical values of the forecast uncertainty that control water transfer are identified under the various WA0. The analysis is illustrated via a case study with a real‐world reservoir in northeastern China. In summary, this study derives hedging rules for water supply reservoir with combined NI and CI to mitigate the risk of both water shortage and water abandonment and provides guidelines to use NI forecasts for water transfer and reservoir release decisions. Plain Language Summary: In this paper, we established a theoretical framework to derive hedging rules for reservoir operation with combined natural inflow (NI) and controlled inflow (CI) through interbasin water transfer. It extends the existing studies on hedging rules, which only considered NI. We particularly address the complexity of CI decisions under NI uncertainty—on the one hand, CI enhances water supply reliability of the recipient reservoir; on the other hand, it can increase the risk of water waste via reservoir spill when the likelihood of a full reservoir is increased. We also addressed the critical values of the forecast uncertainty that control water transfer under the various initial water availability levels. In summary, the theoretical analysis provides guidelines to combine NI and CI to mitigate the risk of both water shortage and water abandonment and provides guidelines to use NI forecasts for water transfer and reservoir release decisions. Key Points: Hedging rules for reservoir operation with combined natural inflow and controlled inflow are derivedWe analyze the effect of forecast uncertainty and water transfer capacity on water transfer and reservoir release decisionsWe identify the uncertainty threshold of natural inflow that controls the decisions [ABSTRACT FROM AUTHOR]

The projection of runoff in a large‐scale basin under climate change and human activities is crucial for the future management of water supplies. The impacts of climate change on runoff projections are associated with large uncertainties. In this study, using the optimization and selection of climate models, as well as an uncertainty analysis via the global Bayesian model averaging (BMA) approach, an ensemble projection framework was established to significantly improve the reliability of runoff projections in the future. The global BMA algorithm proposed in this study is an improvement on the BMA algorithm for large basins and is designed to reflect the differences among various models across the basin. In this algorithm, comprehensive BMA weights are obtained by considering the number (N) of stations in a basin. To verify the feasibility and improvement of this method, the runoff in 2050 and 2070 was projected with data from the Yellow River Basin, China. The runoff and its 90% confidence intervals at the six main stations in the Yellow River Basin were obtained. The increased evapotranspiration will exceed the increase in runoff generated by increase in precipitation in the future. The runoff in the upper and middle reaches of the Yellow River in 2050 and 2070 are projected to be 9% and 7% lower, respectively, than those in the reference period. The ensemble projection method proposed in this paper can be used as a widely applicable process in hydrometeorological ensemble projection and provides a basis for water resource management planning. Key Points: An ensemble projection framework was established to significantly improve the reliability of runoff in the futureThe global BMA algorithm is proposed to obtain comprehensive BMA weights by considering number of stations in a basinThe runoff of the Yellow River Basin in 2050 and 2070 will decrease by 9% and 7%, respectively [ABSTRACT FROM AUTHOR]

This paper presents a new Two Stage Bayesian Stochastic Dynamic Programming (TS-BSDP) model for real time operation of cascaded hydropower systems to handle varying uncertainty of inflow forecasts from Quantitative Precipitation Forecasts. In this model, the inflow forecasts are considered as having increasing uncertainty with extending lead time, thus the forecast horizon is divided into two periods: the inflows in the first period are assumed to be accurate, and the inflows in the second period assumed to be of high uncertainty. Two operation strategies are developed to derive hydropower operation policies for the first and the entire forecast horizon using TS-BSDP. In this paper, the newly developed model is tested on China's Hun River cascade hydropower system and is compared with three popular stochastic dynamic programming models. Comparative results show that the TS-BSDP model exhibits significantly improved system performance in terms of power generation and system reliability due to its explicit and effective utilization of varying degrees of inflow forecast uncertainty. The results also show that the decision strategies should be determined considering the magnitude of uncertainty in inflow forecasts. Further, this study confirms the previous finding that the benefit in hydropower generation gained from the use of a longer horizon of inflow forecasts is diminished due to higher uncertainty and further reveals that the benefit reduction can be substantially mitigated through explicit consideration of varying magnitudes of forecast uncertainties in the decision-making process. [ABSTRACT FROM AUTHOR]

The purpose of this paper is to derive the general optimality conditions of the commonly used operating policies for reservoirs in parallel with joint water demand, which are defined in terms of system-wide release rules and individual reservoir storage balancing functions. Following that, a new set of release rules for individual reservoirs are proposed in analytical forms by considering the optimality conditions for the balance of total water delivery utility and carryover storage value of individual reservoirs. Theoretical analysis indicates that the commonly used operating policies are a special case of the newly derived rules. The derived release rules are then applied to simulating the operation of a parallel reservoir system in northeastern China. Compared to the performance of the commonly used policies, some advantages of the proposed operation rules are illustrated. Most notably, less water shortage occurrence and higher water supply reliability are obtained from the proposed operation rules. [ABSTRACT FROM AUTHOR]

A stochastic programming with recourse model requires the consequences of recourse actions be modeled for all possible realizations of the stochastic variables. Continuous stochastic variables are approximated by scenario trees. This paper evaluates the impact of scenario tree reduction on model performance for hydropower operations and suggests procedures to determine the optimal level of scenario tree reduction. We first establish a stochastic programming model for the optimal operation of a cascaded system of reservoirs for hydropower production. We then use the neural gas method to generate scenario trees and employ a Monte Carlo method to systematically reduce the scenario trees. We conduct in-sample and out-of-sample tests to evaluate the impact of scenario tree reduction on the objective function of the hydropower optimization model. We then apply a statistical hypothesis test to determine the significance of the impact due to scenario tree reduction. We develop a stochastic programming with recourse model and apply it to real-time operation for hydropower production to determine the loss in solution accuracy due to scenario tree reduction. We apply the proposed methodology to the Qingjiang cascade system of reservoirs in China. The results show: (1) the neural gas method preserves the mean value of the original streamflow series but introduces bias to variance, cross variance, and lag-one covariance due to information loss when the original tree is systematically reduced; (2) reducing the scenario number by as much as 40% results in insignificant change in the objective function and solution quality, but significantly reduces computational demand. [ABSTRACT FROM AUTHOR]

This paper addresses how much flood water can be conserved for use after the flood season through the operation of reservoir by taking into account the residual flood control capacity (the difference between flood conveyance capacity and the expected inflow in a lead time). A two-stage model for dynamic control of the flood-limited water level (the maximum allowed water level during the flood season, DC-FLWL) is established considering forecast uncertainty and acceptable flood risk. It is found that DC-FLWL is applicable when the reservoir inflow ranges from small to medium levels of the historical records, while both forecast uncertainty and acceptable risk in the downstream affect the feasible space of DC-FLWL. As forecast uncertainty increases (under a given risk level) or as acceptable risk level decreases (under a given forecast uncertainty level), the minimum required safety margin for flood control increases, and the chance for DC-FLWL decreases. The derived hedging rules from the modeling framework illustrate either the dominant role of water conservation or flood control or the trade-off between the two objectives under different levels of forecast uncertainty and acceptable risk. These rules may provide useful guidelines for conserving water from flood, especially in the area with heavy water stress. The analysis is illustrated via a case study with a real-world reservoir in northeastern China. [ABSTRACT FROM AUTHOR]

The paper evaluates the reform in the irrigation sector in northern China, in particular, the transfer of irrigation management to water user associations or contractors from the village committee. With a set of panel data collected in randomly selected villages in northern China, a fixed effects model at the canal level with instrumental variable estimation is used to control for unobservable heterogeneity and endogeneity problem. The results show that WUAs have increased maintenance expenditures, the timeliness of water deliveries, the percent of irrigated area and the rates of fee collection. There are also improvements in irrigation systems managed by contractors but with magnitudes smaller than in the case of WUAs. WUAs or contracting, however, have limited impacts on water use and crop production. Discussions on reasons for the findings of limited impacts offer some suggestions for the next steps of the irrigation reform and call for continued research efforts to collect more data for further impact evaluations. [ABSTRACT FROM AUTHOR]

The complementary relationship (CR) between actual and potential evaporation has undergone rapid development over the past decades. Commonly, the evaluation of CR models does not comprehensively cover spatial and temporal aspects of model performance; thus, the spatial and temporal accuracy and parameter sensitivity of different CR models remain relatively unknown, especially in ungauged basins. We examine the spatial and temporal performance and parameter sensitivity of four CR models with fixed parameters at the monthly scale over the source region of the Yellow River in China. Additionally, two CR models with distributed parameters were selected as comparisons. Because of the lack of evaporation spatiotemporal "true" values, the corrected water‐balance‐derived evaporation was used as the temporal reference, and the ensemble mean of multiple evaporation products was used as the spatial reference. We find that trade‐offs between the spatial and temporal accuracy of CR models with fixed parameters for basin evaporation should be considered in the application. Although four CR models with fixed parameters exhibited different spatial and temporal performances, their overall performance was generally similar. The parameters of CR models had consistent spatial and temporal sensitivities, and the sensitivity of parameter αe was stronger than that of other parameters. This study demonstrates that CR models with distributed parameters have a promising future in estimating basin or regional evaporation. An integrated spatial and temporal evaluation framework can better distinguish the performance of CR models. The methodologies can be used as a guideline for evaluating and calibrating CR models that match an application's needs. Key Points: Trade‐offs between the spatial and temporal accuracy of complementary relationship (CR) models with fixed parameters for basin evaporation should be consideredThe parameters of CR models had consistent spatial and temporal sensitivities, and the sensitivity of αe is strongerCR models with distributed parameters can significantly improve time series and spatial distribution accuracy [ABSTRACT FROM AUTHOR]

The generalized complementary relationship (CR) of evaporation was investigated in the irrigated winter wheat/summer maize double‐cropping sites in North China Plain (NCP), a region influenced by the East Asian dry winter and wet summer monsoon. Different generalized CRs exist in the winter wheat (dry) and summer maize (wet) seasons with abrupt changes in June and October. The seasonal abrupt change in the CR is attributed to the sudden reversed water vapor transport with the winter and summer monsoon and the wheat/maize rotation with irrigation in the winter wheat season. The separation and abrupt change in the complementary behavior require that two‐stage phased parameters corresponding to the wheat (dry) and maize (wet) seasons should be adopted by the generalized complementary functions for operational evaporation estimation in the irrigated double cropping NCP. Plain Language Summary: The irrigated double cropping North China Plain (NCP) is characterized with rapid winter wheat/summer maize rotation on the land surface and sudden reversed water vapor transport in the upper‐air with East Asian monsoon, coincidentally in June and October. There are corresponding abrupt changes of the complementary relationship, a theory relating changes in landscape evaporation and atmospheric evaporation demand. It reveals the impacts of the processes related to large‐scale changes in atmospheric circulations on the complementary behavior and the requirement of different parameters during the two stages for operational evaporation estimation using the complementary principle in the irrigated double cropping NCP. Key Points: Different generalized complementary relationships (CRs) exist in winter wheat and summer maize seasons with an abrupt changeAbrupt change of the CR is mainly attributed to upper‐air circulations with East Asian monsoonWheat season irrigation contributes to the abrupt change via interrupting the connection between land surface and atmosphere [ABSTRACT FROM AUTHOR]

Terrestrial water storage (TWS) is a critical component for sustainable societal development and ecosystem cycles. The Gravity Recovery and Climate Experiment satellites have tracked changes in global TWS under the combined effects of various factors with unprecedented accuracy since 2002. In this study, we separate the trends in TWS driven by precipitation and non‐precipitation factors for the Chinese mainland from 2003 to 2016 based on the statistical reconstruction method and linear regression and analyze the driving mechanisms combining with multi‐source data. The results show that: (a) during the study period, TWS shows different degrees of increase in most of the Yangtze River basin, the northern part of the Tibetan Plateau, and part of the Heilongjiang Province, while TWS shows a significant decrease in the Tien Shan Mountains, the southeastern part of the Tibetan Plateau, and the North China Plain; (b) precipitation is one of the dominant factors leading to the rise of TWS, and the construction of reservoirs and dams also contributes. In contrast, anthropogenic activities (agricultural irrigation, industrial water use, etc.) and accelerated glacial melting due to global warming are the dominant factors in the decline of water storage; (c) the contribution of long‐term precipitation change to water storage is significantly larger in the northern China region (north of the 800 mm isopleth). This study provides a feasible method for quantifying the contribution of precipitation and non‐precipitation factors to TWS, which is meaningful for understanding the impact of climate change and anthropogenic factors on water resources. Plain Language Summary: Water scarcity is a global crisis for social development in the present and future. As the world's most populous country, China's per capita freshwater resources are far below the global average, making the imbalance between water supply and demand in China even more serious. The study of water storage changes contributes to understanding freshwater resources in China. Here, the precipitation‐ and non‐precipitation‐ driven water storage changes in China are separated based on the statistical reconstruction method and long‐term precipitation data. In addition, satellite gravity, model simulations, and in‐situ data are combined to analyze the driving mechanism in regions that experienced significant changes in water storage during the past decades. Human activities and glacier melting are the dominant factors leading to the water storage deficit. Both precipitation and dam construction contribute to the increase in water storage. This study would provide valuable information for the rational allocation of water resources and coordinated development of the economy and ecology. Key Points: Terrestrial water storage (TWS) trends induced by long‐term precipitation change, precipitation variability, and non‐precipitation factors are separated in ChinaThe TP‐TWS factor is proposed to characterize the impact of long‐term precipitation trends on the TWS trendsThe TWS changes in typical regions are rationalized from the perspective of precipitation and non‐precipitation factors [ABSTRACT FROM AUTHOR]

The extent and mechanisms by which precipitation extremes affect hydrological processes in the critical zone (CZ) of loess possessing a thick unsaturated zone remains poorly understood. To this end, we here employed a coupled liquid‒vapor‒heat‒airflow STEMMUS (simultaneous transfer of energy, mass, and momentum in unsaturated soil) model to investigate the impact of extreme precipitation, at both event and annual scales, on CZ hydrological processes within a semiarid loess site vegetated by apple trees on China's Loess Plateau. At the event scale, the vapor flux was two orders of magnitude lower than the liquid water flux. However, the thermal vapor flux penetrated to depths of 200 cm, whereas the isothermal liquid water flux only infiltrated to 100 cm during the study period, implying that thermal‐gradient‐driven vapor transfer is an important mechanism for deep‐layer recharge (DLR). At the annual scale, the DLR below 200 cm during extremely wet years was 6.5 times larger than that during extremely dry years. In contrast, extreme changes in climate had only limited impacts on evapotranspiration; the difference between extremely wet and extremely dry years averaged only 35 mm, much less than the difference in precipitation, which averaged 310 mm. However, the extremely dry years showed a higher ratio (0.58) between transpiration and evapotranspiration than did the extremely wet years (0.51). The findings reported here improve our understanding of CZ hydrological processes related to precipitation extremes in semiarid loess regions. Key Points: Thermal‐gradient‐driven vapor transfer is an important mechanism for deep‐layer recharge (DLR) in loessThe DLR below 200 cm during extremely wet years was 6.5 times larger than that during extremely dry yearsExtreme changes in precipitation had only limited impacts on evapotranspiration [ABSTRACT FROM AUTHOR]

A new approach combining groundwater storage change (GWSC) derived from Gravity Recovery and Climate Experiment (GRACE) satellite data and baseflow was proposed to estimate groundwater recharge at large spatial scales but a short (monthly) time scale. This method was applied in two typical karst basins of southwest China, the Wujiang River basin (WRB, ∼87,900 km2, ∼70% karstification) and Xijiang River Basin (XRB, ∼360,000 km2, ∼44% karstification). The 2006–2012 monthly baseflow was first separated from in situ streamflow through multivariate regression analysis. Groundwater recharge was estimated using separated baseflow and GWSC estimated from GRACE data and observation‐based groundwater‐level data. The comparison between GRACE‐ and observation‐based recharge in the larger XRB shows better consistency, while that in the smaller WRB shows larger discrepancies. Considering the associated uncertainty (60–93 mm/yr), the 2006–2012 mean recharge in two basins, ranging 199–225 mm/yr (17%–19% of precipitation) is comparable to bulletin‐reported estimates (∼145 mm/yr). Both the degree of karstification and aquifer water table depth influence recharge and discharge processes in karst areas of southwest China. The more karstic WRB exhibits more rapid infiltration (due to more developed permeable epikarst zone) and higher discharge capability (due to more developed underground drainage systems), particularly during dry periods (2009, 2011, and winter). Baseflow exhibits a shorter time lag to recharge and GWSC in the XRB than WRB owing to more quick flow, lower storage with more rapid infiltration, and shorter mean groundwater residence time. Observation‐based monthly recharge reflects different rainfall‐infiltration‐runoff processes under different rainfall intensities, particularly in the WRB. Plain Language Summary: Groundwater recharge is a critical water budget component. This study proposed a new method to estimate groundwater recharge at large spatial scales but on a short time scale. This method combines the groundwater storage changes estimated from gravity satellite data and baseflow (i.e., the main groundwater discharge component) separated from in situ streamflow through multiple regression analysis. We applied this method in two karst basins in southwest China, the Wujiang and Xijiang River Basins. We validated groundwater recharge estimates from gravity satellite data against those only from ground observations. At a larger scale in the Xijiang River Basin, the satellite‐ and ground‐based estimates of groundwater recharge showed better agreement. In the Wujiang River Basin, relatively more rainfall infiltrated into the subsurface to form recharge and more groundwater drained out during dry periods because this basin has a larger proportion of karst areas where the underlying rocks were eroded with large voids and spaces to transfer water. In the Xijiang River Basin, the groundwater storage requires less time to drain out to form baseflow because the flow velocity in this basin is faster. Compared to the gravity satellite data, the ground‐based estimates better reveal the runoff generation mechanism in karst areas. Key Points: A new approach combining Gravity Recovery and Climate Experiment (GRACE) and baseflow is proposed for estimating groundwater recharge at large spatial scales but a short time scaleGRACE‐ and observation‐based recharge estimates are overall consistent, but they indicate different recharge processesGRACE reveals different behaviors of recharge and baseflow during dry periods between two basins with different degrees of karstification [ABSTRACT FROM AUTHOR]

Precipitation intensity–duration–frequency (IDF) curves play a crucial role in the design and planning of urban infrastructure to reduce the risk of urban flooding and rainfall‐triggered landslides. However, changing rainfall characteristics in a warming climate render conventional IDF curves inappropriate due to the statistical assumption of stationarity. In this study, we develop vine copula‐based projections of future IDF curves at sub‐daily to multi‐day time scales with a multi‐model ensemble of five regional climate simulations over China. Stochastic spatiotemporal downscaling of precipitation is achieved to generate extreme precipitation simulations at a high spatial (0.1°) and temporal (3 hourly) resolution. These downscaled simulations are combined by the vine copula to improve the reliability and accuracy of climate‐model‐based IDF curves relative to historical observations. Our findings reveal that climate‐model‐based stochastic downscaling of precipitation reproduces the IDF curves well based on historical observations in China. The vine copula multi‐model ensemble approach outperforms Bayesian model averaging by generating more accurate and reliable IDF curves. The urban areas of 196 Chinese cities are projected to experience an increase in extreme precipitation of up to 30% in intensity and nearly two times the frequency of historical events under a high emission scenario (RCP8.5). The current urban infrastructure of more than half of the 196 cities would thus be inadequate to prevent losses caused by rainfall‐triggered hazards if designed solely based on historical precipitation observations. Compared to the climate‐model‐based IDF curves, we find that statistical IDF curves based on a nonstationary time covariate (i.e., extrapolating historical trends) are likely to underestimate the risk of urban infrastructure failures under a warming climate. This work highlights that urban infrastructure design guidelines in China should be upgraded to adapt existing IDF curves to the changing climate. Key Points: Stochastic spatiotemporal downscaling of precipitation is achieved to derive intensity–duration–frequency (IDF) curves at sub‐daily to multi‐day time scalesA vine copula multi‐model ensemble approach is proposed to improve the accuracy and reliability of climate‐model‐based IDF curvesExtreme precipitation is projected to increase by up to 30% in intensity and nearly two times in frequency over 196 Chinese cities [ABSTRACT FROM AUTHOR]

Evapotranspiration (ET) is a key component of water cycle and is strongly modulated by below‐ground (e.g., dynamic effective rooting depth, Zr) and above‐ground (e.g., LAI and canopy height, CH) vegetation dynamics. Existing studies mainly focus on the effect of above‐ground vegetation dynamics on ET, while it is still unclear how Zr affects ET. Moreover, it is challenging to parameterize Zr dynamically in large‐scale hydrological and biogeochemical models due to data scarcity. Here, we estimate Zr in China from 1982 to 2015 based on Guswa's carbon cost‐benefit model, and update ET partitioning algorithm by replacing CH with Zr in PT‐JPLsm model to form the PT‐JPLzr model. We find that root mean square error (RMSE) between modeled (based on PT‐JPLzr model) and observed ET is 8.25% lower than that of PT‐JPLsm model at 16 in‐situ ET observation sites. Comparing with T/ET from the satellite‐based products, our results highlight the improved performance of PT‐JPLzr (R2 = 0.66) due to incorporated Zr, which is superior to that of PT‐JPLsm model (R2 = 0.62). Transpiration (T) limited by changes in soil moisture (SM) is more sensitive to Zr than CH. Thus, Zr improves the PT‐JPL model performance by affecting the sensitivity of T to soil moisture deficit. Moreover, SM and LAI are main drivers of T/ET spatial variability in drier and wetter regions, respectively. These findings highlight the critical role of Zr in regulating the effects of soil moisture deficit on ET. Key Points: We extend the static plant rooting depth to estimate the dynamic effective rooting depth (Zr) across ChinaZr is incorporated into evapotranspiration (ET) model to simultaneously consider the joint effects of canopy structure and rooting depth on ETTranspiration limited by soil moisture deficit is more sensitive to changes in Zr than canopy height (CH) [ABSTRACT FROM AUTHOR]

Canopy structure alters net precipitation inputs, partly governing the quantity of water recharging soil moisture. Clumped and scattered shrublands are structured with aggregated and isolated canopies, respectively, demonstrating contrasting self‐organized patterns. However, the influence of self‐organization on rain harvesting is largely unknown. Hence, we compared rainfall redistribution patterns of different self‐organized shrubs of Vitex negundo and soil moisture responses during the 2020–2021 rainy seasons on the Loess Plateau of China. Our results indicated that the scattered shrubs harvested more throughfall (85.6% vs. 74.7%) and net precipitation (90.8% vs. 83.8%) than clumped shrubs. Comparatively, stemflow of clumped shrubs was initiated (57.2 vs. 60.4 min) and peaked (198.9 vs. 207.7 min) earlier, ceased later (84.4 vs. 54.5 min), lasted longer (8.9 vs. 8.4 hr), transported more swiftly (397.0 vs. 373.8 mm h−1), and yielded a larger quantity (400.8 vs. 355.1 mL), respectively. This flux was funneled more efficiently (160.1 versus 140.5 fold compared to rain), and was more productive (1.768 vs. 1.346 mm g−1) per unit biomass investment. For both self‐organized patterns, more throughfall led to wetter soils, but more stemflow resulted in quicker response of soil moisture. Comparatively, the top‐layer soil moisture remained more stable post rain under clumped shrubs. Therefore, via rainfall partitioning, the scattered organization was conducive for V. negundo to harvest more rain, but the clumped shrubs harvested rain more efficiently. This might relate to morphological adaptations of shrubs to resist drought and consequent formation and maintenance of self‐organizations at the landscape scale. Plain Language Summary: Rain is intercepted by canopies and redistributed to stemflow (i.e., the rain delivered over woody frames), throughfall (i.e., the rain falling through canopy gaps), and interception loss (i.e., the rain evaporated from canopies). This process is significantly affected by canopy structures, and partly governs the when, where and how of replenishing soil water. The clumped and scattered trees are structured with aggregated and isolated canopies. Although commonly observed, they were rarely compared for different patterns of redistributing rains and recharging soil water. Hence, we measured stemflow, throughfall and soil water of clumped and scattered Vitex negundo, the representative shrub species on the Loess Plateau of China during 2020–2021 rainy seasons. Our results indicated that scattered V. Negundo harvested more throughfall than clumped shrubs, which led to wetter soil beneath isolated canopies. However, clumped V. Negundo harvested more stemflow with quicker starting and peaking, slower ending, and more intense delivering than scattered shrubs, which triggered a quicker response of soil water below aggregated canopies. This might indicate the quantity and efficiency strategies of harvesting rain. In general, via rainfall partitioning, the scattered organization was more conducive for V. Negundo to harvest more rain, but the clumped shrubs harvested rain more efficiently. Key Points: Contrasting self‐organized patterns (clumped vs. scattered) of Vitex negundo reveals different rain harvesting strategiesGreater throughfall and net precipitation occurred beneath scattered shrubs, leading to a larger rain harvest than clumped shrubsHigher stemflow yield and efficiency of denser branches and more closed canopy of clumped shrubs prompted quicker soil moisture responses [ABSTRACT FROM AUTHOR]

Flash drought occurs in a short timescale and has more adverse impacts on society and ecosystems than regular drought. Considering flash drought as a special type of regular drought and identifying it by the rate of the first intensification period might ignore posterior intensification periods. To this end, we introduced the combinatorial decomposable flash drought identification framework (CDFDI) to identify flash drought from the perspective of regarding it as the subphase of drought and explore the characteristics and vegetation response of flash drought in the Loess Plateau (LP), China, where is prone to drought and its ecological environment is very fragile. The neffectiveness of the framework is evaluated through the comparison analysis with a preexisting identification method proposed by Li et al. (2020, https://doi.org/10.5194/essd-12-1217-2020). Results indicate that: (a) the CDFDI framework can better capture flash drought events than the Liu's method in terms of time series and spatial patterns, its rationality was further demonstrated by the obvious latitudinal pattern of correlation between flash drought and normalized difference vegetation index (NDVI); (b) the correlation shows an southeast (positive) and northwest (negative) contrasting pattern in the LP; (c) such contrasting pattern was also found in the response time of GPP, which suggests a more rapid impact on vegetation of flash drought in the arid northwestern LP; and (d) flash drought shows an exacerbating trend across the LP under the background of climate warming and mainly concentrated in the southern part. In general, this study provides new insights on flash drought connotations in a changing environment. Plain Language Summary: Flash drought is an extreme disaster in which drought condition develops rapidly. Identification method considering flash drought as a special type of regular drought and regarding it as an entire drought event has potential limitations. In this study, we develop a new flash drought identification framework that considers it as the subphase of regular drought to fill this gap, and use it to explore the flash drought characteristics and vegetation response in the Loess Plateau (LP). The framework showed a better performance in capturing flash drought compared with the method proposed by Li et al. (2020, https://doi.org/10.5194/essd-12-1217-2020). It can be found that the southern LP is the hotspot of flash drought and large areas of the LP are dominated by summer. Larger exacerbating trend was also observed in the southeastern part which could be related to the warming climate. The correlation between flash drought and normalized difference vegetation index (NDVI) is positive in the southeastern LP and negative in the northwestern LP, and such contrasting pattern was also found in the response time of GPP, where the value is much shorter in the arid northwestern LP. It suggests a more rapid impact on vegetation of flash drought in the area that needs to be alerted. Key Points: An improved framework to identify flash drought considering it as the subphase of drought is proposed and verified in the Loess PlateauThe contrasting pattern was observed in the response of normalized difference vegetation index and gross primary production to flash drought across the Loess PlateauThe southern Loess Plateau is the hotspot of frequent and severe flash drought with an increasing trend [ABSTRACT FROM AUTHOR]

Meteorological and soil drought is disastrous to natural and social systems. It is expected that the occurrence of meteorological and soil drought compound events will become more frequent and extreme in the future. The propagation process of meteorological drought to soil drought plays an essential role in the occurrence of drought. However, the drought propagation process is not clear, especially the quantification of drought characteristic propagation process is rarely realized. This study constructs drought intensity propagation index (DIP) and drought duration propagation index (DDP), and then puts forward drought propagation partition in China. Furthermore, this study preliminarily discussed the driving factors of drought propagation and the relationship between land‐atmosphere interaction and main drought propagation partition (arid type drought propagation area, peer‐to‐peer type drought propagation area, and humid type drought propagation area). The results show that DDP and DIP are significantly negatively correlated in China from 1981 to 2020. From southeast to northwest, the propagation of drought intensity from meteorological drought to soil drought gradually increased and the propagation of drought duration gradually decreased. We further concluded that 68% of regions with similar intensity and duration of meteorological drought and soil drought (peer‐to‐peer type drought propagation area) are concentrated in semiarid and dry subhumid, which is exactly the area with strong land‐atmosphere interaction. In all land use types, grasslands are most prone to peer‐to‐peer propagation of drought. This study further reveals the law and mechanism of drought propagation and provides a new idea and attempt to clarify the drought propagation process. Key Points: Drought intensity propagation index and drought duration propagation index for quantifying drought propagation characteristics are proposedDividing drought propagation partitionDiscovery of drought propagation close peer‐to‐peer in areas with strongest land‐atmosphere coupling [ABSTRACT FROM AUTHOR]

To support the implementation of China's carbon neutrality target, this study assesses the impact of future climate change on hydroelectricity production (HP) in China. We developed a hydropower database for mainland China, which covers 92% of the installed capacity indicated by national statistics (352 GW in 2018). An HP model was applied to simulate monthly HP by 3,521 conventional plants at different levels of global warming (1°, 1.5°, 2°, and 3°C above preindustrial) using an ensemble of monthly streamflow time series computed by five global hydrological models (GHMs), where each global hydrological model was driven by four global climate models (GCMs). With a probability of 80%, limiting global warming to 1.5°C will constrain changes of national HP to the range of −1.4%–2.5% as compared to near‐present climate (1°C). In case of a 3°C world, this range more than doubles. HP is projected to likely increase in the two provinces Sichuan and Yunnan that encompass more than half of China's installed capacity. However, with a probability of about 10%, HP may decrease in a 3°C world by more than 15% in five provinces, and 40% of the installed capacity may suffer from an HP decrease of more than 5%. The uncertainty of projected HP changes is mainly due to the GCMs. We found that relative changes in annual HP and streamflow often differ in magnitude and direction as HP depends on seasonal streamflow patterns and installed capacities that determine the maximum flow through the turbines. Key Points: A new hydropower database for mainland China was compliedSignificant reductions due to climate change cannot be precluded for 40% of the installed capacityChanges in annual hydroelectricity production and streamflow often differ in magnitude and direction [ABSTRACT FROM AUTHOR]

Quantification of the dependence characteristics in hydrological time series is essential for understanding hydrological variability and for managing water resources. However, how determining a suitable model for describing the dependent components is still a challenge. In this article, we proposed a correlation coefficient‐based information criterion (CCIC) to determine the optimal model order of regression‐based models for quantifying the dependence characteristics in hydrological time series. CCIC was developed by combining the index of correlation coefficient and the information entropy index. The former was used to reflect the fitting error of the model and the latter was used as a penalty term to reflect the model uncertainty. They have similar roles as the two terms of the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) used commonly in hydrology. Results of Monte‐Carlo experiments verified higher and more stable accuracy of determining the true model order by CCIC than by AIC and BIC. Moreover, results indicated that the mean value of a time series had a big impact on the accuracy of CCIC. If the mean value of a time series was far from its initial value, the estimation of CCIC would have a big bias, causing low accuracy in the determination of suitable model order. The application of CCIC to annual precipitation from 520 stations in China further confirmed its advantages over AIC and BIC and indicated the significant short‐term dependence characteristics of annual precipitation in the Yangtze River basin. The proposed CCIC approach has potential for wider use in hydrometeorology. Key Points: Correlation coefficient‐based information criterion (CCIC) was proposed to determine the optimal order of regression models for quantifying dependence characteristics in hydrologic time seriesMC experiments verified higher and more stable accuracy of determining the true model order by CCIC than by Akaike Information Criterion and Bayesian Information Criterion used commonlyMean value of a time series had a big impact on the accuracy of CCIC for determining optimal order for regression models [ABSTRACT FROM AUTHOR]

In recent studies, deep reinforcement learning (RL) methods have been used for the real‐time control of urban drainage systems (UDSs). However, the training process of an RL agent is computationally expensive since collecting the training data requires multiple simulations of a UDS model. An effective solution to this issue is to replace the original UDS model with an emulator, a simpler model which emulates the specified output of the system. However, the adaptability and interpretability of the existing UDS emulators still require further improvements. To address these challenges, a new emulator with increased adaptability and interpretability, called Koopman emulator, was developed by using the Koopman operator and a deep learning technique in this study. Then, the Koopman emulator was used to replace the UDS model in RL to establish an emulator‐based RL framework for fast and efficient training. This framework was used to train emulator‐based RL agents for flooding and overflow mitigation in a UDS located in Eastern China. Baseline emulators were also employed for comparison. All the methods were evaluated through simulations. The following results were obtained: (a) The Koopman emulator demonstrated a better emulation performance than the baseline emulators. (b) The RL agents based on the Koopman emulator achieved a similar control effect with a faster training process and more efficient data usage compared with the RL agents based on the storm water management model. (c) The uncertainty analysis showed that the RL agents based on the Koopman emulator exhibited acceptable robustness when facing different rainfall events and input perturbations. Key Points: Koopman emulator demonstrates better performance and better interpretability in nonlinear emulationThe emulator‐based reinforcement learning (RL) framework achieves similar control effect with faster training process and more efficient data usageThe RL agents based on Koopman emulator exhibited acceptable robustness when facing different rainfall events and input perturbations [ABSTRACT FROM AUTHOR]

We present a Generative Adversarial Network (GAN)‐based 3D reservoir simulation framework, GANSim‐3D, where the generator is progressively trained to capture geological patterns and relationships between various input conditioning data and output earth models, and is thus able to directly produce multiple 3D realistic and conditional earth models from given conditioning data. Conditioning data can include 3D sparse well facies data, probability maps, and global features, such as facies proportion. The generator only includes 3D convolutional layers, and once trained on a data set consisting of small‐size data cubes, it can be used for geomodeling of 3D reservoirs of large arbitrary sizes by simply extending the inputs. To illustrate how GANSim‐3D is practically used and to verify GANSim‐3D, a field karst cave reservoir in Tahe area of China is used as an example. The 3D well facies data and 3D probability map of caves obtained from geophysical interpretation are taken as conditioning data. First, we create training, validation, and test datasets consisting of 64 × 64 × 64‐size 3D cave facies models integrating field geological patterns, 3D well facies data, and 3D probability maps. Then, the 3D generator is trained and evaluated with various metrics. Next, we apply the pretrained generator for conditional geomodeling of two field cave reservoirs of size 64 × 64 × 64 and 336 × 256 × 96, respectively. The produced reservoir realizations prove to be diverse, consistent with field geological patterns and field conditioning data, and robust to noise in the 3D probability maps. Each realization with 336 × 256 × 96 cells only takes 0.988 s using 1 GPU. Key Points: Generative Adversarial Network (GAN)‐based 3D reservoir simulation framework (GANSim‐3D) produces multiple 3D realistic conditional reservoir models of large arbitrary sizes from various conditioning dataApplication in field karst cave reservoirs shows robust performance of GANSim‐3DGeomodeling with GANSim‐3D is quite fast: <1 s for one realization with 336 × 256 × 96 cells [ABSTRACT FROM AUTHOR]

Real‐time control (RTC) of urban drainage systems (UDS) has been proved an efficient tool in combined sewer overflow (CSO) and flooding mitigation. Recently, new RTC approaches based on reinforcement learning (RL) were developed for flooding mitigation in stormwater systems. While these studies have made contributions to enable an improved urban water management, they are insufficient to allow for deeply understanding of the effectiveness of different RLs in UDS. Meanwhile, the risk of handing over the control process to a RL agent is still unavoidable because of the fluctuations of RLs' output and the unknown consequences of implementing RLs control strategy. This study conducted four tasks to address these problems. First, five RTC systems based on five individual RLs were designed to distinguish different RLs' performance in the context of UDS. Then, an independent security system based on SWMM was provided to forecast and evaluate the consequence of RL control strategy. After that, an innovative hybrid RTC system, called Voting, was developed by coupling multiple RLs and the independent security system through a model predictive control framework to avoid the fluctuations of RLs' output. Finally, the robustness of the RL agents was validated using uncertainty analysis. All the RLs were evaluated through simulation based on a Storm Water Management Model of a UDS located in Eastern China. According to the results, (a) different RLs show promise in CSO and flooding mitigation; (b) Voting selects a relatively reliable and optimal control trajectory compared with any single RL agents; (c) the performances of RL agents have certain robustness when facing different rainfall events and imperfect input. Key Points: Different reinforcement learnings (RLs) show promise in CSO and flooding mitigationVoting selects a relatively more reliable and more optimal control trajectory compared with any single RL agentsRLs' performance is influenced by rainfall events and imperfect input, but they still have certain robustness [ABSTRACT FROM AUTHOR]

Groundwater use underpins much economic production. The unsustainable use of groundwater threatens environmental flows in surface waters, sustainable development, and future food security. The connection between agricultural trade and groundwater depletion has been recently highlighted, but how groundwater depletion supports the production of industrial and tertiary goods, trade, and consumption remains less well understood. Here, we present the first analysis of groundwater use and depletion embedded in the complete supply chain of China (including primary, secondary, and tertiary products). We use a multiregion input‐output analysis coupled with the high‐resolution groundwater use modeling to track groundwater depletion from production to end consumer. Our modeling results show that groundwater depletion occurred primarily in water scarce North China for agricultural production, but the depleted resource was then incorporated throughout the supply chain and dispersed across Chinese and international consumers. ∼64 billion m3 yr−1 (±1 billion m3 yr−1) groundwater was depleted in China, in which more than a half was from Xinjiang, Hebei, Henan and Heilongjiang Provinces. Approximately 40% of the groundwater depletion can be traced to interprovincial transfer (21 billion m3 yr−1) and export (4.8 billion m3 yr−1). The hot spots for final consumption of groundwater depletion were major cities in both North and South China. Importantly, over 60% of the groundwater depletion was embodied in industrial and tertiary products for final consumption, highlighting the importance of tracing groundwater through the full economy. Groundwater depletion represents a long‐term risk to supply chains, and policy‐makers can use this understanding to sustainably manage groundwater and diversify supply chains. Plain Language Summary: Groundwater is a critical resource that is under increased pressure due to unsustainable exploitation. Groundwater resources underpin both national supply chains and global trade, but how groundwater supports the full supply chain of a nation remains unresolved. This study traces the groundwater depletion incorporated in the complete supply chain (including industrial and tertiary products, not just primary goods) of China through interprovincial transfers and international exports. The results show how groundwater depletion, which is concentrated in North China, is distributed to domestic and international consumers of final goods. International importers receive China's groundwater in the receipt of industrial products. We highlight that groundwater depletion represents a long‐term risk to supply chains. Key Points: Groundwater depletion is traced using a multi‐region input‐output analysis coupled with the high‐resolution groundwater use modelingAbout 21 billion m3 yr−1 groundwater depletion, primarily from water scarce Northern‐China, is virtually transferred domesticallyGroundwater depletion supporting export to the Rest of the World is estimated to be 4.8 billion m3 yr−1 [ABSTRACT FROM AUTHOR]

Afforestation to control soil erosion has been implemented throughout China over the past few decades. The long‐term hydrological effects, such as total water yield and baseflow, of this large‐scale anthropogenic activity remain unclear. Using six decades of hydrologic observations and remote sensing data, we explore the hydrological responses to forest expansion in four basins with contrasting climates across China. No significant change in runoff was found for the period 1970–2012 for the cold and dry Hailar River Basin in northeastern China. However, both forest expansion and reduced precipitation contributed to the runoff reduction after afforestation since the late 1990s. Similarly, afforestation and drying climate since the mid‐1990s induced a significant decrease in runoff for the Weihe River Basin in semi‐arid northwestern China. In contrast, the two wet basins in the humid southern China, Ganjiang River Basin and Dongjiang River Basin, showed insignificant changes in total runoff during their study periods. However, the baseflow in the winter dry seasons in these two watersheds significantly increased since the 1950s. Our results highlight the long‐term variable effects of forest expansion and local climatic variability on basin hydrology in different climatic regions. This study suggests that landuse change in the humid study watersheds did not cause dramatic change in river flow and that region‐specific afforestation policy should be considered to deal with forestation‐water quantity trade‐off. Conclusions from this study can help improve decision‐making for ecological restoration policies and water resource management in China and other countries where intensive afforestation efforts are taking place. Key Points: Both precipitation and afforestation explained reduction in runoff for dry basinsForest recovery did not result in significant changes in total runoff for two wet basinsDry season baseflow in two wet basins had an upward trend [ABSTRACT FROM AUTHOR]

Despite a multitude of models predicting sediment transport dynamics in open‐channel flow, self‐organized vertical density stratification that dampens flow turbulence due to the interaction between fluid and sediment has not been robustly validated with field observations from natural rivers. Turbulence‐suppressing density stratification can develop in channels with low channel‐bed slope and high sediment concentration. As the Yellow River, China, maintains one of the highest sediment loads in the world for a low sloping system, this location is ideal for documenting particle and fluid interactions that give rise to density stratification. Herein, we present analyses from a study conducted over a range of discharge conditions (e.g., low flow, rising limb, and flood peak) from a lower reach of the Yellow River, whereby water samples were collected at targeted depths to measure sediment concentration and, simultaneously, velocity measurements were collected throughout the flow depth. Importantly, sediment concentration varied by an order of magnitude between base and flood flows. By comparing measured concentration and velocity profiles to predictive models, we show that the magnitude of density stratification increases with sediment concentration. Furthermore, a steady‐state calculation of sediment suspension is used to determine that sediment diffusivity increases with grain size. Finally, we calculate concentration and velocity profiles, showing that steady‐state sediment suspensions are reliably predicted over a range of stratification conditions larger than had been previously documented in natural river flows. We determine that the magnitude of density stratification can be predicted by a function considering an entrainment parameter, sediment concentration, and bed slope. Key Points: Detailed concentration profiles reveal grain‐size specific density stratificationSediment diffusivity effects are overwhelmed by density stratification at high concentrationsFine washload sediment (<25 μm) is vertically stratified in the Yellow River [ABSTRACT FROM AUTHOR]

Forecast errors of multiple information sources of a cascade hydropower system cause risks of water and energy supply in real‐time operation. Mechanisms minimizing multiple risks with budgeted cost under oscillations of multiple forecast uncertainties through robust operation are not yet well‐investigated. This study proposed a multiobjective robust optimization (RO) and decision‐making framework comprising series of models for risk analysis, robust control, and decision making. The risk analysis model identifies and analyzes dependent risks that stem from forecast errors of supply and demand‐side information by Copula functions. A RO model was established for minimizing risk probabilities, vulnerabilities of ecological protection, water and energy supply, as well as revenue loss from energy. Thereafter, a multiattribute decision‐making model was incorporated for determining the compromise solution from competing non‐dominated solutions. The proposed framework was applied to the cascade hydropower system of the Xiluodu, Xiangjiaba, Three Gorges project and Gezhouba reservoirs on the Yangtze River, China. The principal findings were as follows. (a) Addressing the singular forecast uncertainty of streamflow is inadequate to obtain robust solutions. (b) In a normal year, RO can reduce the average conditional value‐at‐risk of ecological water shortfall, consumptive water shortfall, and energy shortfall by 25.1%, 35.3%, and 16.5% than that of chance‐constrained programming. (c) Multiple risks are dispersed through risk avoidance and hedging that reduces outflow variance and homogenizes mean outflow, lowering down efficiencies of energy production and water usage as a tradeoff. The proposed model framework could be applied to inform operation of cascade hydropower systems. Key Points: Holistic multiobjective robust optimization model framework for reservoir operation under supply and demand uncertainties was developedConflicts of reducing risk probabilities, vulnerabilities, and revenue loss for ecological protection, water supply, and energy were quantifiedMechanisms of tradeoffs in multiobjective robustness with energy production through risk hedging and avoiding were discovered [ABSTRACT FROM AUTHOR]

Much dissolved inorganic carbon (DIC) may exist in lacustrine carbonates in the terminal lakes of closed basins in arid areas, and inorganic carbon sinks in arid regions are an important hypothesis of global carbon cycle research in recent years. Previous studies have made significant progress in estimating the inorganic carbon content. However, the source of inorganic carbon sinks remains unclear. Here, we choose four typical arid closed basins in North China to investigate the provenance of inorganic carbon sinks. The results show that CO2 absorbed by saline‐alkaline soil is the main source of groundwater DIC in the study area, with a contribution rate of approximately 50%. Forty‐three groundwater radiocarbon ages from arid closed basins indicate that the DIC in groundwater is transported along the direction of water flow and deposited in terminal lakes, forming inorganic carbon sinks. Our study suggests that the total inorganic carbon contents of Zhuye Lake, Yanchi Lake, Huahai Lake, and Qinghai Lake in the Holocene are 0.318 Pg, 0.003 Pg, 0.160 Pg, and 3.289 Pg, respectively. The results demonstrate that inorganic carbon sinks are much larger than lake organic carbon storage, thereby serving as an important carbon cycle factor in arid regions. Plain Language Summary: As observation methods and experimental technologies have advanced, research on the global carbon cycle has penetrated all links of the Earth's surface system. The long‐neglected arid zone ecosystem exerts a significant impact on the interannual changes in global terrestrial carbon sinks. Despite this, research on the source of inorganic carbon sinks in arid areas is still insufficient. Therefore, age measurements and isotopic analyses are conducted in groundwater, river water and sediments in four arid closed basins of North China, and the main sources of inorganic carbon sinks in arid closed basins are explored. The results show that groundwater dissolved inorganic carbon is mainly derived from the CO2 absorbed by the saline‐alkaline soil through river leakage and irrigation and finally reaches the terminal lakes. Taken together, soil CO2 is one of the leading sources of inorganic carbon sinks in closed basins. Key Points: CO2 absorbed by saline‐alkaline soil is the main source of groundwater dissolved inorganic carbon (DIC) with a contribution rate of about 50%The DIC formed by the carbon absorption of saline‐alkaline soils becomes an inorganic carbon sinkThe rate of Holocene inorganic carbon sinks from lacustrine carbonates is much higher than that of organic carbon accumulation [ABSTRACT FROM AUTHOR]

Facing the dual pressures of carbon mitigation and trade barrier, it is desired that variations of water footprints (WFs) with the related policy interferences be investigated in Yangtze River Economic Belt (YREB). In this study, a factorial equilibrium WFs model is developed to (a) tackle the interactive effects (on blue‐ and gray‐WFs) of different policy alternatives presented as multiple levels of carbon tax and import tariff; (b) explore the variations of blue‐ and gray‐WFs in specific socio‐economic sectors under multiple scenarios of the dual pressures; and (c) investigate the provincial WFs from the perspective of commodity consumption. It is found that increased import tariffs can boost the WFs of primary energy and resource‐conversion sectors, and can promote inter‐sectoral virtual‐water exchanges; carbon tax can suppress the WFs for most of the sectors, and can result in entirely declined industrial production. Moreover, carbon tax can lead to reduced water productivity in YREB, and thus exacerbate water shortage. Through this research, desired policies for water‐footprint management policies could be identified with maximized socio‐economic and environmental benefits. Key Points: A factorial equilibrium water footprint (WF) model for exploring the contributions of multiple policy alternatives on changes in WFs is proposedHigh carbon tax may sacrifice China's economy; trade barrier between China and the US may lead to negative impacts on China's food securityCarbon tax and trade barrier would reduce the blue WFs in the upper reach and would have opposite effects on the gray WFs in the lower reach [ABSTRACT FROM AUTHOR]

China has experienced significant vegetation greening over recent decades. Vegetation greening may regulate the effective vegetation available water capacity (EVAWC), which determines the root zone water availability for evapotranspiration. EVAWC is one of the most important factors controlling vegetation growth and ecosystem resilience. Despite its importance, quantifying EVAWC at large scale is challenging due to limited observations of soil properties and insufficient understanding of underlying physical processes. Thus, to what extent changes in EVAWC are driven by vegetation and climate change is still unclear. Here, we establish a physics‐based, data‐driven stochastic approach to examine the spatiotemporal variability of the EVAWC and its potential drivers over China, with the consideration of vegetation and climate dynamics. The major strength of our approach is that it does not require observations related to soil properties as inputs, but only needs a few simple hydrometeorological and remotely sensed vegetation parameters that can be easily obtained. On average, we find that vegetation changes tend to increase EVAWC over 71.0 ± 3.9% of the total land area in China, while climate variabilities tend to reduce EVAWC over 73.7 ± 9.5% of the land area. More importantly, the potential increase in EVAWC resulting from vegetation changes is offset by climate variability. As a result, the changes in EVAWC over recent decades are generally dominated by climate variabilities (relative contribution of 82.7 ± 9.6%) rather than vegetation changes, especially in humid and semihumid areas. Our results suggest that vegetation greening in China has subordinate impacts on EVAWC, and ecosystem resilience is mainly dominated by the local climate. Key Points: A stochastic approach is developed to examine changes in effective vegetation available water capacity (EVAWC) over ChinaThis approach does not require observations related to soil properties as inputs, which are difficult to be obtained and have uncertaintiesChanges in EVAWC are dominated by climate variabilities (relative contribution 82.7 ± 9.6%) rather than vegetation changes [ABSTRACT FROM AUTHOR]

Streamflow simulations at daily time steps are vital to water resources management, especially in arid regions. Previously, data‐driven models have been used as an effective tool for daily streamflow simulation. However, the accuracy of conventional data‐driven approaches is affected by the temporal autocorrelation of daily streamflow, especially in irrigated watersheds where the persistence of saturated flows dominates irrigation seasons. This study presents a Stepwise Clustered Regression Tree Ensemble (SCRTE) to address the streamflow autocorrelation. With the provision of a state‐of‐the‐art data‐driven model Stepwise Cluster Analysis (SCA), the SCRTE enables both single‐ and multi‐output settings (i.e., model predictand can be either a scalar or a vector), which can thus address interactions among streamflow values over multiple consecutive days. The autocorrelation effect of daily streamflow is evaluated based on single‐ and multi‐output SCA ensembles, which can then be aggregated according to their performance for various streamflow quantile ranges. To facilitate the irrigation scheduling decision‐making under rigorous transboundary water regulations, the SCRTE is applied to three interconnected watersheds with mixed land use, located in a floodplain of the Yellow River basin in China. The results show that the SCRTE outperforms seven well‐known benchmark models across seven evaluation metrics. Our findings reveal that the SCRTE can reflect the varying effects of autocorrelation over different streamflow quantile ranges, thereby improving the streamflow simulation. The multi‐output SCA ensembles are more capable of addressing the medium flows, while the single‐output one can better simulate the low and high flows. Key Points: A regression tree ensemble is proposed to address the temporal autocorrelation of daily streamflowA Bayesian model average with a stratified sampling ensemble strategy is provided to improve the simulation accuracy for medium flowsAutocorrelation of daily streamflow is particularly important for establishing reliable irrigation‐discharge relationships [ABSTRACT FROM AUTHOR]

Land Surface Model (LSM) predictions are affected by unknown parameters, uncertain meteorological forcings, inaccurate initial conditions, and incomplete process representation. While parameter optimization (PO) algorithms have been used to estimate the model parameter for best model performance, the use of data assimilation (DA) has been receiving increased prominence to characterize the uncertainties in such models. However, it is still an open question that how DA techniques can be used together with PO algorithms to enhance the performance of LSMs for different applications. In this study, a combined framework composed of the PO algorithm—Adaptive Surrogate Modeling Based Optimization (ASMO) and the DA algorithm—Evolutionary Particle Filter with Markov Chain Monte Carlo (EPFM), is used to improve the soil moisture (SM) estimates of Community Land Model (CLM) across the Tibetan Plateau in China. The SM data from the Soil Moisture Active Passive satellite are used as the benchmark data for model improvement. To demonstrate the usefulness and effectiveness of the proposed approach, the simulated SM values are validated against the SM observations collected from the in situ networks in the Tibetan Plateau, China. The findings revealed that the joint application of ASMO and EPFM algorithms results in more accurate and reliable SM estimates of CLM compared to when they are solely employed. This study suggests that the combined optimization‐assimilation framework can be utilized for improving the predictive skill of the other large‐scale complex LSMs while accounting for the uncertainties associated with both model parameters and state variables. Key Points: A combined framework by jointly using parameter optimization and data assimilation algorithms is proposedA new objective function is introduced in PO to make the optimization spatially effectiveThe value of assimilation of satellite soil moisture data into the Community Land Model is investigated [ABSTRACT FROM AUTHOR]

Sustainable stormwater management (SSM) has gained global interests to solve both on‐site and off‐site downstream water problems. However, current structural SSM practices are often limited to small scales and their impacts on river water quality have not been fully understood. In this study, we have quantified watershed responses to SSM via multiyear monitoring and integrated modeling in an urbanizing watershed in China, while five scenarios (baseline, current SSM, improved SSM, future traditional urban development, and future improved SSM scenarios) were comparatively analyzed. The results showed that water quality of ammonia nitrogen (NH3‐N), total phosphorus (TP), chemical oxygen demand (COD), and suspended solid (SS) would deteriorate by both traditional urban development and current SSM design. Compared to other pollutants, NH3‐N and TP showed greatest deterioration by current SSM due to disproportionate management of on‐site runoff quantity‐related and quality‐related problems. Comparatively, the improved SSM design, which incorporated watershed responses into consideration, would reduce current concentrations of NH3‐N, TP, COD, and SS by 0–53.85%, 8.16–38.84%, 23.07–61.03%, and 68.82–77.82%, respectively; and improve water quality by 60–80% in future SMM scenario. Priority management and stricter objectives should be given to those urbanizing regions at the watershed scale. The results of this study should be used as references for future SSM design as well as water resource protection. Plain Language Summary: Sustainable stormwater management (SSM) is vital for urban water resources and water quality protection. However, watersheds serve as the end of the SSM system, the related water quality benefits from SSM have not been fully quantified. To address this problem, five scenarios were evaluated via multiyear monitoring and integrated modeling in an urbanizing watershed. Water bodies in the urbanizing areas are more vulnerable, which would multiply several times in pollutant concentrations by traditional urbanization. However, unexpected deterioration of river quality occurred by current SSM, although local runoff and pollutant loads were controlled. The improved SSM was then given by incorporating watershed response into SSM design to handle water problems at different scales. The findings and approach could undoubtedly facilitate the SSM development and watershed management, and could be applied in other cities and countries because urban expansion has become a global issue. Key Points: Unexpected river quality deterioration was found in current sustainable stormwater management (SSM) due to the neglect of relative reduction proportions in runoff and pollutant loadsVulnerable water bodies were found in urbanizing period, during which NH3‐N and TP were more sensitive to urbanization and SSMImproved SSM design was put forward by considering watershed responses through an integrated model [ABSTRACT FROM AUTHOR]

Climate elasticity of runoff is an important indicator for evaluating the effects of climate change on runoff. Consequently, this paper proposes an analytical derivation of climate elasticity. Based on the mean annual water-energy balance equation, two dimensionless numbers (the elasticities of runoff to precipitation and potential evaporation) were derived. Combining the first-order differential of the Penman equation, the elasticities of runoff to precipitation, net radiation, air temperature, wind speed, and relative humidity were derived to separate the contributions of different climatic variables. The case study was carried out in the Futuo River catchment in the Hai River basin, as well as in 89 catchments of the Hai River and the Yellow River basins of China. Based on the mean annual of climatic variables, the climate elasticity in the Futuo River basin was estimated as follows: precipitation elasticity ϵP} = 2.4$, net radiation elasticity ϵRn = - 0.8, air temperature elasticity ϵT} = -0.05°C-1 wind speed elasticity ϵU} = -0.3, and relative humidity elasticity ϵRH0.8. In this catchment, precipitation decrease was mainly responsible for runoff decline, and wind speed decline had the second greatest effect on runoff. In the 89 catchments of the Hai River and the Yellow River basins of China, climate elasticity was estimated as follows: ϵPranging from 1.6 to 3.9, ϵRn ranging from -1.9 to -0.3, ϵET ranging from -0.11 to -0.02°C-1, ϵU ranging from -0.8 to -0.1, and ϵRH ranging from 0.2 to 1.9. Additional analysis shows that climate elasticity was sensitive to catchment characteristics. [ABSTRACT FROM AUTHOR]

Irrigation is vital for the sustainability of water‐limited agricultural areas. However, higher irrigation efficiency (IE) rarely reduces irrigation water consumption, a paradox well noted in the real world. Resolving this paradox requires rigorous assessment of basin‐scale IE, which remains rare. This study holistically analyzed basin‐scale IE using the Zhangye Basin (ZB), a typical agricultural area in arid northwestern China, as the testbed. A new basin‐scale IE index was proposed that factors in irrigation return flow and groundwater's direct contributions to crop evapotranspiration. A novel approach was also developed to calculate this new index based on integrated ecohydrological modeling. The major study findings for the ZB include the following. First, return flow accounts for approximately 13% of irrigation water, which creates a difference of 0.1 between the new basin‐scale IE index and a traditional field‐scale index. Second, although basin‐scale IE has notable interannual fluctuation, its multiyear average shows stability, which may be dependent on the basin's physical characteristics. Third, basin‐scale IE reflects large spatial heterogeneity and is influenced by the intensity of surface water‐groundwater interaction. Finally, under the projected future climate, return flow in the ZB will be enhanced by 3.2% per decade, and basin‐scale IE shows an increasing trend. Overall, rigorous evaluation of basin‐scale IE is critical to making heterogeneous policies and developing adaptive management strategies under the changing climate. The integrated modeling approach developed in this study provides a methodological foundation for overcoming misunderstandings about the IE paradox and benefits the reform of the current IE policy agenda. Key Points: A basin‐scale irrigation efficiency index was proposed to consider irrigation return flow and crop ET directly from groundwaterA novel approach was developed to assess basin‐scale irrigation efficiency based on integrated ecohydrological modelingThis study provides a methodological foundation for overcoming misunderstandings about the paradox of irrigation efficiency [ABSTRACT FROM AUTHOR]

Characterizing the upper tail of flood peak distributions remains a challenge due to the elusive nature of extreme floods. China has experienced some of the most extreme floods on Earth and is therefore an important setting for extreme flood research. Here, we examine the upper tail of flood peaks over China based on a comprehensive flood dataset that integrates systematic observations from 1759 stream gaging stations and 14,779 historical flood surveys. Floods with large upper tail ratios (i.e., flood peak discharge normalized by sample 10‐yr flood) cluster spatially in the windward regions of mountainous terrain with basin mean annual rainfall of around 550 mm. Large upper tail ratio floods are associated with a mixture of rainfall generation processes. These processes are characterized by anomalous moisture transport (e.g., zonal water vapor pathways) and/or synoptic configurations (e.g., blocking), with respect to flood‐producing storms for annual flood peaks. Downscaling simulations with the Weather Research and Forecasting model for the 1939, 1963, and 2016 floods over the North China Plain region highlight the role of interactions between complex terrain and large‐scale environment in defining the world envelope curve of flood peaks. The upper‐tail floods demonstrate a combination of ingredients producing extreme rainfall. Our results provide important insights to extreme floods in China and a frame of reference for assessing the largest floods on Earth. Key Points: We examine the upper tail of flood peaks based on the most comprehensive dataset of flood records to date across ChinaComplex terrain and mean annual rainfall are two key elements in dictating the distribution of upper tail ratiosAnomalous moisture transport provides the fundamental distinction between large upper tail ratio floods and annual flood peaks [ABSTRACT FROM AUTHOR]

Many deep aquifers overlain by barrier formations in the continental US are used as geological repositories for wastewaters coproduced from hydrocarbon exploration. This practice is to protect shallow groundwater following the US Environmental Protection Agency's regulation. Implicit in such practice is the assumption that deep aquifers overlain by mudstones or shales are confined so that the injected fluids will not migrate upward to contaminate shallow groundwater. However, no systematic test of this hypothesis has been made. Here we invert the groundwater response to both the M2 and the O1 tides and to the barometric pressure across a large (2.046 × 106 km2) geologic regime, the North China Platform, to systematically evaluate the hydraulic parameters as functions of depth and time without a priori assumption. Our result, the first of such inversion, shows no depth dependence of aquifer confinement to a depth of 3,400 m and that deep confined aquifers overlain by barrier formations may become leaky after distant earthquakes. It suggests that monitoring of aquifer confinement may be needed to ensure if the targeted deep aquifer for wastewater injection is really confined. The results may be timely and of global significance because the practice of hydrofracking for natural gas and the deep injection for the disposal of the coproduced wastewaters in the US may soon be adopted by other countries, such as China. Plain Language Summary: A great amount of wastewaters coproduced from hydrocarbon exploration in the mid‐continental US has been injected into deep aquifers. Such practice has been based upon an implicit assumption that deep aquifers confined by aquitards are fully confined so that the injected toxic water will not leak back to contaminate shallow groundwater or the environment. This assumption, however, has never been carefully tested. Here we test this hypothesis using groundwater data from wells over a large geological regime, the North China Platform, as a function of depth and time. Our results show no clear dependence of aquifer confinement on depth, that significant leakage occurs even in the deepest studied aquifer (>3400 m) confined by thick clay/mud aquitards, and that an initially confined aquifer became may leaky after a distant earthquake. Because the practice of hydrofracking for natural gas and deep injection for the disposal of the coproduced wastewaters may soon be adopted by other countries, such as China, our study is timely and may be of global significance. Key Points: Systematic evaluation of the confinement of aquifers in the North China Platform at different depths down to 3.4 kmNovel revealing most of the studied aquifers leak, and has no dependence on aquifer depth or thickness of the overlain geologic barrierBoth the M2 and O1 tidal waves are used in the inversion of aquifer parameters without a prior assumption [ABSTRACT FROM AUTHOR]

We present a multi‐objective, multi‐stage stochastic programming with recourse model for reservoir management and operation, where we use utility theory to select the best compromise solution from the Pareto front. A multi‐stage streamflow scenario tree is generated first by the neural gas method. Then the Pareto front at each stage is produced by a modified constrained NSGA‐II. A single best compromise solution on the Pareto front must be selected for the immediate stage and the model moves forward one stage and is re‐optimized over a moving planning horizon of fixed duration. A key contribution of this study is the selection of the best compromise solution, which is achieved by a proposed linear spline utility function allied with regression. Our proposed utility function has the following advantages: (a) It satisfies the law of diminishing marginal rate of substitution, (b) it does not rely on the pre‐specified weight or goal, and (c) it selects the best compromise solution that is likely to fall in the "knee regions" of the Pareto front. We apply the proposed optimization model to the Three Gorges Reservoir in China. The two conflicting objectives are (a) maximizing the total expected energy output in the planning horizon, and (b) maximizing the average expected ecological benefits in the planning horizon. The results show that the proposed model produces the optimal water release policy successfully under different hydrological scenarios, considering both the inflow uncertainty and the tradeoff between the two conflicting objectives. Key Points: Presents a multi‐objective, multi‐stage stochastic programming with recourse model for reservoir operationsDevelops a linear spline utility function for selecting the most preferred solution on the Pareto frontEvaluates the performance of the proposed optimization framework with a real‐world case study [ABSTRACT FROM AUTHOR]

Significant attention has recently been paid to deep learning as a method for improved catchment modeling. Compared with process‐based models, deep learning is often criticized for its lack of interpretability. One solution is to combine a process‐based hydrological model with a residual error model based on deep learning to give full scope to their respective advantages. In classical residual error models, Bayesian inference via Markov chain Monte Carlo (MCMC) is commonly used to provide an estimation of the uncertainty. However, deep neural networks tend to have excessively large numbers of parameters, making MCMC an unsuitable approach. Here, we introduce an alternative to Bayesian MCMC sampling called stochastic variational inference (SVI) which has recently been developed for Bayesian deep learning in Natural Language Processing. We implement SVI in a Long Short‐Term Memory (LSTM) network and construct residual error models in process‐based hydrological models. This approach is examined in the contrasting geographical and climatic characteristics of two catchments from China, the Tangnaihai catchment and the Shiquan catchment. Compared with the Bayesian linear regression model, the Bayesian LSTM provides better uncertainty estimates. Specifically, the proposed method improves the Continuous Ranked Probability Score (CRPS) by over 10% in both two catchments. In the Tangnaihai catchment, it provides more than 10% narrower uncertainty intervals in terms of Sharpness with slightly superior Reliability. In the Shiquan catchment, it provides comparable uncertainty intervals with better Reliability. Further, our study highlights the scalability of SVI to high‐dimensional parameter spaces in hydrological applications (e.g., distributed hydrological models, groundwater models). Key Points: Residual error models are constructed to take advantages of process‐based hydrological models and Bayesian deep learningBayesian Long Short‐Term Memory provides better streamflow uncertainty estimates than a Bayesian linear regression modelStochastic variational inference as an alternative to Markov chain Monte Carlo is viable to estimate tens of thousands of parameters [ABSTRACT FROM AUTHOR]

Monitoring and prediction of agricultural drought are paramount to food security at the global and regional scales, particularly under the influence of climate change and anthropogenic activities. Soil moisture is an effective indicator for monitoring and characterizing agricultural drought. Soil moisture (agricultural drought) is mainly affected by precipitation (meteorological drought) and temperature (hot conditions). Owing to the flexibility of vine copulas in handling multidimensional variables by decomposing them into pair copula constructions (PCCs), we propose a novel drought prediction method using three predictors, namely antecedent meteorological drought, previous hot conditions, and persistent agricultural drought, based on the conditional distributions of C‐vine copulas in a four‐dimensional scenario. The proposed model was applied to agricultural drought (characterized by the standardized soil moisture index (SSI)) prediction with 1–2‐months lead time for the summer season (i.e., August at a 6‐months timescale) in China. Taking two severe agricultural drought events that occurred in many regions across China in August of 2006 and 2014 as validation cases, the SSI predictions with 1–2‐months lead time using the conditional C‐vine copulas model were found to be generally consistent with the corresponding historical SSI observations in most parts of China. Performance evaluation using the Nash‐Sutcliffe efficiency (NSE), coefficient of determination (R2), and F1 score (F1S) for different climate regions also indicated that this model provided a reliable prediction of agricultural drought for most areas of China. The outcome of this study can serve as a guidance for drought prediction, early warning, and drought mitigation. Key Points: A novel drought prediction model is proposed via conditional vine copulas for ChinaWe develop a detailed approach for agricultural drought prediction based on the vine‐copulaGood prediction performance is presented for different climate regions of China [ABSTRACT FROM AUTHOR]

Understanding nitrate migration through the deep vadose zone is essential for aquifer vulnerability assessments. The effect of variability of physical properties of the deep vadose zone on nitrate transport has been scarcely explored. Recently, deep nitrate storage profiles were determined in the vadose zone of the Loess Plateau of China. Using these observations along with measured soil properties, this study investigates the effect of loess vertical heterogeneity on water movement and nitrate transport through the deep vadose zone. Models of different complexity were established and calibrated. First, a simple piston flow and nitrate mass balance approach was calibrated to the observed nitrate storage. The results indicate that the total nitrate storage is estimated well, while the estimation of the distribution of nitrate is relatively poor. Subsequently, Richards' equation and the Advection‐Dispersion equation were evaluated. Three different conceptualizations of the numerical models were calibrated against deep vadose zone nitrate and water content observations: (1) one‐layer model assuming homogenous loess vadose zone; (2) a model that considers a hydraulic conductivity (Ks) decay function and (3) a model where the Miller‐Miller scaling factors are prescribed to account for changes of the hydraulic functions with depth. Accounting for the vertical Ks decay in the numerical models improved water flow performances. The study reveals the adequacy of implementing water flow and nitrate transport numerical models together with a simple representation of the vertical loess variability, for simulating nitrate migration in loess deep vadose zone environments. Plain Language Summary: Enhanced concentration of nitrate in groundwater is a global problem. The source of such nitrate is commonly linked to agricultural practices, in particular the use of fertilizers. The intensive development of China in the last few decades has put at risk many groundwater systems, such as the unconfined aquifer of the Loess Plateau of China (LPC). Understanding the fate and travel times of nitrate through the LPC vadose zone before its arrival at the water table would help sustain groundwater systems and identify areas requiring changes in land use management. The current study examines various modeling approaches for nitrate transport estimations in the LPC vadose zone. Four modeling approaches with different complexities are tested: A simple mass‐balance model and more sophisticated numerical models that include different methods to represent the vertical variability of loess physical properties. The study reveals that water flow and nitrate transport numerical models together with a simple representation of the vertical loess variability is the preferable approach for describing nitrate transport in loess environments. Key Points: Examination of how complex a model needs to be for the simulation of water and nitrate movement in deep unsaturated loessThe performance of a combined Richards' equation and Advection‐Dispersion equation approach is superior to a simplistic piston flow modelUsing a Ks decay function to represent loess vertical variability in numerical models improves unsaturated water flow predictions [ABSTRACT FROM AUTHOR]

In this study, an innovative MODIS fractional snow cover (SCF) data assimilation (DA) prototype framework that invokes machine learning (ML) techniques and Common land model (CoLM) is proposed to improve the estimation of the snow depth (SD) and the SCF. To validate our new framework, we analyzed two snow seasons from 2013 to 2015 at 46 stations in Northern Xinjiang in China. We developed 12 SCF DA schemes that represent different DA methods (direct insertion (DI) and Ensemble Kalman Filter (EnKF)), observational data (original data and gap‐filled MODIS SCF data), and observation operators (five new snow depletion curves (SDCs) defined using traditional multivariate nonlinear regression and four ML methods). While improving the frequency of the SCF observations in the DI‐based DA scheme only resulted in a marginal improvement in the snow estimates, by adding new SDCs fitted by ML techniques (e.g., deep belief network), and the gap‐filled MODIS SCF data to the EnKF‐based DA scheme, we were able to reduce model structural uncertainties of CoLM and achieve marked improvement in the accuracy of the snow estimates (RMSE = 5.92 cm, mean bias error = −1.94 cm, and average degree of improvement = 32.18% for SD estimates and RMSE = 15. 79%, mean bias error = −1.21%, and average degree of improvement = 47.95% for SCF estimates). Our results demonstrate the feasibility of improving snow estimates by combining the ML techniques with physically based snowpack model in a SCF DA framework. Key Points: An innovative MODIS fractional snow cover assimilation framework invokes machine learning techniques and Common Land Model is proposedFive new snow depletion curves (SDCs) are defined by using multiple nonlinear regression and four machine learning techniquesNew SDCs functions as both the improved parameterization scheme of Common Land Model and the observation operator of assimilation frame [ABSTRACT FROM AUTHOR]

There is a drastic geographic imbalance in available global streamflow gauge and catchment property data, with additional large variations in data characteristics. As a result, models calibrated in one region cannot normally be migrated to another without significant modifications. Currently in these regions, non‐transferable machine learning models are habitually trained over small local data sets. Here we show that transfer learning (TL), in the senses of weight initialization and weight freezing, allows long short‐term memory (LSTM) streamflow models that were pretrained over the conterminous United States (CONUS, the source data set) to be transferred to catchments on other continents (the target regions), without the need for extensive catchment attributes available at the target location. We demonstrate this possibility for regions where data are dense (664 basins in Great Britain), moderately dense (49 basins in central Chile), and scarce with only remotely sensed attributes available (5 basins in China). In both China and Chile, the TL models showed significantly elevated performance compared to locally trained models using all basins. The benefits of TL increased with the amount of available data in the source data set, and seemed to be more pronounced with greater physiographic diversity. The benefits from TL were greater than from pretraining LSTM using the outputs from an uncalibrated hydrologic model. These results suggest hydrologic data around the world have commonalities which could be leveraged by deep learning, and synergies can be had with a simple modification of the current workflows, greatly expanding the reach of existing big data. Finally, this work diversified existing global streamflow benchmarks. Plain Language Summary: We introduced a method to utilize available big data to better start and warm up a machine learning streamflow model that is later fine‐tuned for prediction in basins on other continents (Asia, South America and Europe). This procedure noticeably improved streamflow volume prediction for different scenarios with varying amounts of data in the target basins (in terms of time period, length of collected data, and number of basins having data). This allows thousands of basins across the world with only a few years' worth of streamflow observations to benefit from improved modeling and accuracy resulting from the use of deep learning. Key Points: Basins around the world can be better modeled by applying transfer learning (TL) to a deep network trained in the US, and tuning it locallyThe benefits of TL increased with the amount and diversity of the source data, and were larger than from pretraining with a hydrologic modelThis work greatly expands the reach of deep learning, adds to the value of existing big data, and calls for synergy of global data sets [ABSTRACT FROM AUTHOR]

Since 1999, the Loess Plateau, China, has undergone one of the world's largest revegetation programs (Grain for Green Project, GfGP). Revegetation has profound impacts on hydrological cycle and water balance, especially in arid and semi‐arid areas. As an essential component of water balance, the long‐term change in water storage (ΔS) is generally treated as zero. However, it remains unclear how to define the time scale of "long‐term," especially over regions undergoing dramatic vegetation restoration. In this study, we quantify ΔS over the Loess Plateau from 1982 to 2009 from a water balance perspective, with a particular focus on the impact of vegetation on ΔS at different time scales. Results show that a 44.78% increase in vegetation coverage could lead to a 37.72% decrease in ΔS. Moreover, the better the vegetation growth, the lower the ratio of ΔS to precipitation (ΔS/P) in the Loess Plateau. ΔS/P achieved a local minimum at the 4‐ to 5‐year scale and reached a plateau after 10 years. We further use the Budyko framework to verify the above conclusions. We find that errors of simulated streamflow over 11/16 of catchments become stable after 4–5 years, and all 16 catchments become stable after 10 years. Ignoring ΔS at a 10‐year scale has the least influence on water budget closure. These findings highlight the importance of considering time scales when ignoring ΔS in the water balance analysis, especially when the length of measurement is limited. Key Points: We use simulated evapotranspiration (ET) from the PT‐JPL model, observed streamflow and precipitation to quantify ΔS from 1982 to 2009 on 16 catchments over the Loess Plateau, ChinaWe find that a 44.78% increase in vegetation coverage could reduce ΔS by 37.72%Neglecting ΔS on a time scale of about 10 years would have the least impact on the streamflow error simulated by the Budyko framework in the Loess Plateau [ABSTRACT FROM AUTHOR]

Ground rupture due to groundwater pumping is a major environmental hazard accompanying land subsidence in some areas. Ground ruptures usually develop when a significant compaction affects a sedimentary sequence with peculiar geological conditions, for example, the presence of a shallow bedrock with buried ridges. A laboratory test was developed with the aim at improving our understanding of the mechanisms responsible for rupture generation in this particular hydrogeological setting, typical, for instance, of the Guangming village, China. A 0.8‐m high concrete prism, representing a rocky ridge, was placed in a 4.0‐m long, 1.8‐m wide, and 1.4‐m high box and buried by alluvial material. The box was saturated and then drained, with the formation of a main crack above the prism‐shaped ridge. The measured water content, vertical displacements, strains, rupture initiation, and growth are analyzed through an original coupled 3D‐numerical model, simulating the variably saturated groundwater flow in a deformable and fractured porous medium. To our knowledge, this is the first application of such a kind of model to a lab‐scale experiment. Despite the uncertainty on the material parameters, the numerical model allows satisfactorily reproducing the observed groundwater flow, land subsidence, and rupture features (depth and width). The rupture dynamics is captured in part as well, despite the employment of a simple Mohr–Coulomb failure criterion and although other forces (e.g., electrochemical) may likely play a role at the metric scale of the lab test. The modeling outcomes provide a clear view on how the rupture develops from the surface and propagates downward. Key Points: A large lab experiment was carried out to reproduce rupture development due to aquifer exploitation above a bedrock ridgeA novel coupled 3D model of variably saturated flow in deformable and fractured porous media was developed and applied to the lab experimentThe numerical model allows to describe the brittle rupture initiation and growth over time [ABSTRACT FROM AUTHOR]

Lake Hulun is the fifth‐largest lake in China, playing a substantial role in maintaining the balance of the grassland ecosystem of the Mongolia Plateau, which is a crucial ecological barrier in North China. To better understand the changing characteristics of Lake Hulun and the driving mechanisms, it is necessary to investigate the water storage changes of Lake Hulun on extended timescales. The main objective of this study is to reconstruct the water storage time series of Lake Hulun over the past century. We employed a machine learning approach termed the extreme gradient boosting tree (XGBoost) to reconstruct the water storage changes over a one‐century timescale based on the generated bathymetry and satellite altimetry data and investigated the relationships with hydrological and climatic variables in long term. Results show that the water storage changes from 1961 to 2019 were featured by four fluctuation phases, with the highest water storage observed in 1991 (14.02 Gt) and the lowest point in 2012 (5.18 Gt). The century‐scale reconstruction result reveals that the water storage of Lake Hulun reached the highest point in the 1960s within the period of 1910–2019. The lowest stage occurred in the sub‐period of the 1930s–1940s, which was even lower than the alerted shrinkage stage in 2012. The predictive model results indicate the effective performance of the XGBoost model in reconstructing century‐scale water storage variations, with the mean absolute error of 0.68, normalized root mean square error of 0.11, Nash–Sutcliffe efficiency of 0.97, and correlation coefficient of 0.94. The annual fluctuations of water storage were mostly affected by precipitation, followed by vapor pressure, temperature, potential evapotranspiration, and wet day frequency. The dominating characteristics of different variables vary evidently with different sub‐periods. The atmospheric circulations of the Arctic Oscillation, El Nino Southern Oscillation, Pacific Decadal Oscillation, and North Atlantic Oscillation have tight associations with the water storage variations of Lake Hulun, which change with different study periods. Plain Language Sumamry: To better understand the changing characteristics of Lake Hulun and the driving mechanisms, it is necessary to investigate the water storage changes of Lake Hulun on extended timescales (e.g., century timescale). The main objective of this study is to reconstruct the water storage time series of Lake Hulun over the past one century. We employed a machine learning approach termed the extreme gradient boosting tree (XGBoost) to reconstruct the one‐century water storage changes and investigated the relationships with hydrological and climatic variables in long term. The century‐scale reconstruction result reveals that the water storage reached the highest point in the 1960s within the period 1910‐2019. The lowest stage occurred in the 1930s‐1940s, which was even lower than the alarted shrinkage stage in 2012. The annual fluctuations of water storage were mostly affected by precipitation, followed by vapor pressure, wet day frequency, potential evapotranspiration, and temperature. However, the dominating characteristics of different variables vary evidently with different sub‐periods. This study is expected to provide an efficient technical solution of reconstructing long‐term lake water storage records and to advance our scientific understanding of the characteristics of lake water balances in response to climate change and variability in the Mongolia Plateau. Key Points: The complete bathymetry map of Lake Hulun was constructed based multi‐source remote sensing dataA machine learning approach was employed to reconstruct the century‐scale water storage of Lake HulunThe potential links of one‐century variations of lake water storage with climatic variables and atmospheric circulations [ABSTRACT FROM AUTHOR]

Irrigation is the greatest human interference with the terrestrial water cycle. Detailed knowledge on irrigation is required to better manage water resources and to increase water use efficiency (WUE). This study applies a framework to quantify net irrigation at monthly timescale at a spatial resolution of 1 km2 providing high spatial and temporal detail for regional water resources management. The study is conducted in the Haihe River Basin (HRB) in China encompassing the North China Plain (NCP), a global hot spot of groundwater depletion. Net irrigation is estimated based on the systematic evapotranspiration (ET) residuals between a remote sensing‐based model and a hydrologic model that does not include an irrigation scheme. The results suggest an average annual net irrigation of 126 mm yr−1 (15.2 km3 yr−1) for NCP and 108 mm yr−1 (18.6 km3 yr−1) for HRB. It is found that net irrigation can be estimated with higher fidelity for winter crops than for summer crops. The simulated water balance for NCP is evaluated with Gravity Recovery and Climate Experiment (GRACE) data, and the net irrigation estimates can close the water balance gap. Annual winter wheat classifications reveal an increasing crop area with a trend of 2,200 km2 yr−1. This trend is not accompanied by a likewise increasing trend in irrigation water use, which suggests an increased WUE in the NCP, which is further supported by net primary productivity data. The proposed framework has potential to be transferred to other regions and support decision makers to support sustainable water management. Plain Language Summary: The irrigation of agricultural fields is taking place at unsustainable rates in many regions of the world. Despite the fact that irrigation is the largest anthropogenic impact on the natural water cycle, there exists limited knowledge of the applied irrigation amounts. This study applies an approach to estimate net irrigation, that is, the evaporative loss of irrigation water. The approach is applied on the Haihe River Basin in North‐Eastern China. For the estimation of net irrigation, two sources of evapotranspiration (ET) are considered. First, baseline ET is obtained from a rainfed hydrologic model without irrigation. Second, ET is obtained from a satellite remote sensing model, which represents rainfed and irrigated ET. We study the ET differences of the two sources to derive net irrigation amounts at monthly timescale at 1 km2 spatial resolution. Our analysis suggest an average annual net irrigation of 108 mm yr−1 (18.6 km3 yr−1). The results are evaluated against annual winter wheat classification maps as well as satellite‐based net primary productivity data and total water storage data (Gravity Recovery and Climate Experiment, GRACE). Our results indicate an increasing water use efficiency as a result of promoting water savings in the agricultural sector. Key Points: Net irrigation can be quantified by evapotranspiration residuals derived from hydrologic models and satellite remote sensingEstimation of winter wheat irrigation has lower uncertainty than estimation of irrigation of summer cropsIrrigation water use efficiency has improved in the North Chin Plain in the period 2002–2016 [ABSTRACT FROM AUTHOR]