Reclaimed water replenishment plays a crucial role in sustaining urban river systems in arid regions where natural water resources are scarce. However, its ecological impacts, particularly on microbial community diversity and structure, remain insufficiently understood, with significant knowledge gaps regarding how reclaimed water replenishment influences these dynamics. Here, we quantified changes in water quality and multi-trophic community structure along natural (NR), reclaimed (RR), and mixed (NRR) sections of the Xiaohei River, an arid urban river in northern China. The results indicate that reclaimed water discharge altered the physicochemical conditions of the river systems, particularly in terms of nutrient enrichment, with total nitrogen and total phosphorus concentrations increasing on average by 67% and 84%, respectively. After reclaimed water replenishment, significant shifts in community composition and diversity were observed, with the mean richness of algae and metazoans increasing by 20.2% and 29.7%, respectively, while that of bacteria and protozoa decreased by 19.6% and 4.1%. Nitrogen and phosphorus nutrients, together with hydrodynamic conditions, emerged as the primary drivers of biodiversity variation, with particularly strong associations observed for bacterial, algal, and protozoan communities. Chlorophyta and Rotifera displayed contrasting responses to reclaimed water replenishment, with Chlorophyta exhibiting a monotonic linear response and Rotifera showing a nonlinear (hump-shaped) relationship with physicochemical gradients, underscoring differential responses across trophic levels. This study improves understanding about the ecological effects of reclaimed water replenishment on aquatic ecosystems. It provides a scientific basis for nutrient control and ecological monitoring to support coordinated water allocation and ecological conservation in arid urban river systems.

Study region: Changchun City in North China. Study focus: Water level prediction within sewer networks plays a crucial role in urban water system real-time control. With advanced automated monitoring technologies, the generation of vast amounts of observational data makes it possible to predict water levels using deep learning models. However, a knowledge gap remains regarding how input feature dimensionality and temporal sequence length influence performances of such models. To address this issue, we developed an explainable deep learning framework for predicting the influent water level at the terminal wastewater treatment plant of a typical sewer system by integrating Long Short-Term Memory (LSTM) with SHapley Additive exPlanations (SHAP) for interpretable analysis. New hydrological insights: Utilizing monitored precipitation data and upstream pipeline water level time-series data as input yields satisfactory one-step-ahead prediction performance. Contrary to expectations, incorporating additional upstream sensors did not improve model accuracy. Model performance improved progressively with longer input sequences, reaching optimal performance at 48-time steps before deteriorating with further temporal extension. SHAP analysis revealed that manual control operations exerted stronger influence on model outputs than the spatial proximity of monitoring sites to prediction points. Furthermore, sensor data anomalies and deficiencies in monitoring network configurations could be identified, confirming its utility for diagnosing behavioral patterns in modeled sewer network. These findings provided new hydrological insights and practical guidance for sewer system real-time control.

Understanding the temporal variation in streamflow in the Lancang–Mekong River and its driving mechanism is essential for water resource management of this important international river. In this study, streamflow at the Chiang Saen gauging station was simulated using a long short-term memory (LSTM) model driven by satellite-based Multi-Source Weighted-Ensemble Precipitation (MSWEP) and Multi-Source Weather (MSWX) datasets, with the aim of quantifying the contributions of climate change and human activities to streamflow variations. A key contribution of this work lies in the use of LSTM to reproduce naturalized streamflow conditions—using only climate inputs—thereby providing a data-driven alternative to conventional process-based modeling approaches in this data-scarce basin. The monthly precipitation and temperature data of Chiang Saen station from 1979 to 1991 are used for model training and validation. The natural streamflow of Chiang Saen station from 1992 to 2021 is reconstructed based on the trained model. The results show that the annual average precipitation of the basin from 1979 to 2021 only exhibits a statistically insignificant decreasing trend, while the annual average temperature shows a statistically significant upward trend, and the inter-annual variation in the annual average streamflow shows a non-significant downward trend. Periodic analysis shows that the main periodicity of precipitation, temperature, and streamflow data is 12 months, following annual periodicity in climate. LSTM simulations demonstrate high accuracy in predicting the streamflow in T month based on the MSWEP precipitation and MSWX temperature data in T-2, T-1, and T months. On an annual scale, the streamflow in the changing period (1992–2021) decreases by only 4.6% compared with the reference period (1979–1991). In spring, the streamflow in the changing period is 30.6% higher than that of the reference period, and climate change and human activities contribute 40.8% and 59.2%, respectively. Increases in streamflow (3.4%) are also detected in the winter, with human activity as the dominant contributing factor. For the summer, the streamflow in the changing period is −8.2% lower than that in the reference period, with a greater contribution from human activities (68.7%) than climate change (31.3%). The streamflow in autumn of the changing period is −12.1% lower than that in the reference period, with a greater contribution from human activities (90.2%) than climate change (9.8%). In general, the findings of this study indicate that the driving mechanisms behind streamflow changes at Chiang Saen are complex at different temporal scales, and they provide valuable insights for improving our understanding of hydrological changes within the Lancang–Mekong River Basin.

Abstract. Multiyear droughts (MYDs) are recognized as severe drought events, with especially profound impacts on both human activities and ecosystems. However, the optimal rainfall replenishment timing (toptimal) for MYDs mitigation remains insufficiently understood. With that in mind, we conducted a retrospective analysis of historical MYDs based on the Palmer Drought Severity Index (PDSI) in China during 1961–2020, and the calibration period was set to 1961-1990. We performed a series of numerical experiments involving precipitation gradient increases for 351 selected MYDs, distributed across 199 grids (2°×2°), from 1991 to 2020, and developed a drought mitigation quantitative model (DMQM). In addition, a key coefficient (k) derived from DMQM was defined to quantify the mitigation efficiency, and toptimal was then identified as the timing corresponding to the maximum k (kmax). Overall, drought severity exhibits a nonlinear response to increased precipitation. kmax occurred most frequently in the first month of drought onset (t1), accounting for 58.79 % of all grids, while the second (t2) and third (t3) months were also non-negligible, accounting for 22.11% and 11.06 %, respectively. Compared to the humid river basins in southern China, the arid and semi-arid northern regions had a higher probability for k at t2 or t3 to exceed k at t1. Drought duration (DD) was identified as a key factor, as longer DD was associated with a greater likelihood of t2 or t3 being the toptimal, evidenced by R2 values of 0.526 and 0.578, respectively. These findings contribute to ensuring timely and regionally appropriate MYD mitigation strategies and interventions.

Satellite and reanalysis-based precipitation products have played a crucial role in addressing the challenges associated with limited ground-based observational data. These products are widely utilized in hydrometeorological research, particularly in data-scarce regions like the Qinghai–Tibetan Plateau (QTP). This study proposed an ensemble streamflow simulation method using remote sensing precipitation data as input. By employing a 1D Convolutional Neural Networks (1D CNN), streamflow simulations from multiple models are integrated and a Shapley Additive exPlanations (SHAP) interpretability analysis was conducted to examine the contributions of individual models on ensemble streamflow simulation. The method is demonstrated using GPM IMERG (Global Precipitation Measurement Integrated Multi-satellite Retrievals) remote sensing precipitation data for streamflow estimation in the upstream region of the Ganzi gauging station in the Yalong River basin of QTP for the period from 2010 to 2019. Streamflow simulations were carried out using models with diverse structures, including the physically based BTOPMC (Block-wise use of TOPMODEL) and two machine learning models, i.e., Random Forest (RF) and Long Short-Term Memory Neural Networks (LSTM). Furthermore, ensemble simulations were compared: the Simple Average Method (SAM), Weighted Average Method (WAM), and the proposed 1D CNN method. The results revealed that, for the hydrological simulation of each individual models, the Kling–Gupta Efficiency (KGE) values during the validation period were 0.66 for BTOPMC, 0.71 for RF, and 0.74 for LSTM. Among the ensemble approaches, the validation period KGE values for SAM, WAM, and the 1D CNN-based nonlinear method were 0.74, 0.73, and 0.82, respectively, indicating that the nonlinear 1D CNN approach achieved the highest accuracy. The SHAP-based interpretability analysis further demonstrated that RF made the most significant contribution to the ensemble simulation, while LSTM contributed the least. These findings highlight that the proposed 1D CNN ensemble simulation framework has great potential to improve streamflow estimations using remote sensing precipitation data as input and may provide new insight into how deep learning methods advance the application of remote sensing in hydrological research.

Land use, as an integrated representation of natural conditions and human activities, significantly impacts river water quality. Understanding the spatial and temporal variability of these influences offers valuable insights for improving water quality through the implementation of best management practices. This study examined the impact of land use on river water quality in the Dahei River Basin, a typical mountain-to-plain basin located in the arid region of northern China, which is also the last first-order tributary of Upper Yellow River. Hierarchical clustering analysis was employed to analyze the spatial distribution characteristics of river water quality and redundancy analysis was used to explore the impacts of land use on water quality in upstream buffer zones with radii from 500 m to 14,000 m. The results indicate that river water quality conditions in the mountainous region are much better than in the plain region. In both the dry and wet seasons, land use significantly affects water quality variation, particularly at the 8000 m buffer zone, although the mechanisms differ. In the wet season, the non-point source pollution from storm runoff erosion dominates the positive correlations between water pollution levels and the areas of cropland and urban regions, while for the dry season such positive correlations may come from elevated soil electrolyte levels due to groundwater irrigation and point source pollution from urban activities. For land use types that show a negative correlation with water pollutant levels, the stronger correlation observed in grasslands compared to forests region may be attributed to grasslands' better adaptation to arid conditions. The findings from this study enhance our understanding of the spatiotemporal variations in land use impacts on river water quality and can provide guidance for land use planning at the basin scale in arid regions.

In recent years, population growth and the continuous expansion of agricultural land, non-point source (NPS) pollution has gradually become the main cause of deteriorating water quality in the aquatic environment, particularly in the Yangtze River Basin of China, a highly economically developed region. The landform of middle and upper regions of the Yangtze River is mainly hilly and mountainous, with a large elevation difference resulting soil erosion and NPS pollution. Therefore, the Linjiang River basin (LJRB), a typical hilly area in the upper reaches of the Yangtze River, was selected as the study area. In this study, the spatiotemporal variations of NPS pollution loads including total nitrogen (TN) and total phosphorus (TP) in the LJRB during the period 2014 to 2021 were firstly investigated by using the SWAT model. Additionally, a framework was developed to assess the spatiotemporal variations of NPS pollution risk in 2030 and 2050, providing a novel tool for future land-use planning, which coupled SWAT with the PLUS model. The average values of R2 and NSE for runoff were up to 0.91 and 0.82, respectively, and those for water quality were 0.84 and 0.79, respectively. The results indicated the high accuracy and reliability of the SWAT model simulation outcomes after refinement. In the LJRB, the contribution of TN and TP to water pollution showed a fluctuating upward trend over eight years. With TN reaching 5.98 mg/L in 2017 and TP rising to 1.21 mg/L in 2021, both pollutants ranked as the top contributors to water pollution, making them critical targets in regional pollution control efforts. Regarding seasonal distribution, the crop growing season (April to June) was identified as the critical period for controlling NPS pollution. The high-risk regions were distributed mostly in the cultivated land, which showed that fertilization application was the main cause of NPS pollution. The predictive results for future NPS pollution under 3 scenarios reveal that the NPS pollution load under ecological protection scenario (S3) was the minimum. S3 scenario considers both urban development and environmental protection, and is the best land use scenario setting for the management of the study area. The results obtained in this study have practical significance for reducing the NPS pollution loads in the LJRB and controlling the environmental pollution in the Yangtze River. In addition, this study provides an effective approach to assess spatiotemporal variations of NPS pollution risk caused by both historical and potential future land-use change and to support future land-use planning for effective NPS pollution control.

Spatial resolution of topography data significantly impacts computational time of lake hydrodynamic modelling. This study proposes a calibration tool to examine impacts of topography data resolution on simulation uncertainty, evolving from the Generalized Likelihood Uncertainty Analysis framework. Using the EFDC hydrodynamic model, BaiYangDian Lake in North China was simulated at three resolutions: 200, 500, and 1000 m. The first two models show similar accuracy, outperforming the 1000-m model. The parameter space constrained by water level observations and the simulation uncertainties in water level, water age, and velocity from 500-m model closely resembled those from 200-m model, while requiring only 16.7% of the latter's computational time, indicating a feasible spatial resolution range where model performance matches the high-resolution model but with significantly less computational time. The study highlights the importance of calibration with multiple observations and demonstrates potentials of the proposed tool to identify effects of model settings on simulation uncertainty.

Understanding the mechanisms driving streamflow changes is crucial for effective water resource management in arid regions. However, basin-scale assessments based on modeling approaches are often constrained by data scarcity and simulation uncertainty. This study explored the relative contributions of climate change and human activities to streamflow changes in the Dahei River Basin in the upstream region of the Yellow River, China, using three different machine learning models (MLMs) to reconstruct natural streamflow from 1980 to 2020 at eight streamflow stations. MLMs are constrained by their reliance on the quality, quantity, and representativeness of the data, which can introduce uncertainties in simulations. To address this limitation, in this study, assessment uncertainties were evaluated on the basis of the differences among MLM simulations. The Mann‒Kendall test indicated that while precipitation remains unchanged, temperature and streamflow show significant increasing and decreasing trends, respectively, except at Xierdaohe station, where streamflow increased. On the basis of monthly data, continuous wavelet transform analysis revealed 1-year periodicity for the three hydrometeorological variables, with the periodicity of streamflow vanishing recently. At seven stations, streamflow decreased after the abrupt change, primarily due to human activities, with relative contributions ranging from 58.7 to 119.8

Poyang lake (PYL), as China's largest freshwater lake, has been suffering from increasing total phosphorus (TP) pollution associated with rapid basinal socio-economic development. However, anthropogenic phosphorus stressors were rarely examined in PYL basin due to its large-scale and complex river-lake connection water system, hampering phosphorus pollution control efforts. In this study, water pollution stress from multiple anthropogenic activities is quantitatively examined in PYL basin based on a newly developed framework coupling grey water footprint (GWF) analysis with the SPARROW model. Results show that the phosphorus source-sink process in PYL basin has been well simulated by SPARROW model quantifying an overall TP delivery rate of 0.39, with catchments closer to PYL showing higher delivery rate of phosphorus (up to over 0.8). The GWF analysis demonstrates that anthropogenic phosphorus sources have imposed much higher pollution stress on PYL than its inflowing rivers, with twelve catchments nearby PYL identified as critical source areas of TP contamination. Agricultural farming, livestock & poultry production, and urban household are recognized as the dominant anthropogenic stressors burdening water environment of PYL. Based on these, policy recommendations are provided for advancing control of the phosphorus pollution stressors. The methodology is effective in refined examination of water pollution sources, which is expected to be applied in other watersheds providing informative diagnosis of water issues especially in lakes.

Changes in vegetation growth are important indicators of health of terrestrial ecosystems. Located in China's semi-arid and semi-humid transitional zone, the Fenhe River Basin is the second largest tributary of the Yellow River, which is vital for soil and water conservation in the Loess Plateau. In this study, based on the Normalized Difference Vegetation Index (NDVI) and meteorological data during 1982-2015, we evaluated the spatial-temporal distribution and changes in the vegetation greenness and meteorological variables in the Fenhe River Basin before and after the Grain for Green Program (GGP), in order to reveal the driving factors of changes in vegetation growth, and then quantify their relative contributions. The results show that since the GGP being implemented in 1999, the vegetation coverage in the Fenhe River Basin has significantly improved. Climate change and human activities are two important factors affecting the vegetation greenness detected from NDVI and the combined effects of the two factors play a leading role in NDVI change. After the implementation of GGP, the influence of climate on NDVI change has weakened, and the area proportion of human activities improving vegetation growth has raised significantly. This study is valuable for providing knowledge about the dynamic changes of vegetation in the Fenhe River Basin under the influence of climate change and human activities, and provides decision support information for vegetation restoration and sustainable development in the Fenhe River Basin.

Increasingly intensive crop farming has been aggravating global water pollution. It is significant to retrospectively investigate world's crop production induced water pollution stress, facilitating sustainable agriculture in the future. In this study, nearly 60-yr grey water footprint (GWF) dynamics of global crop farming are examined at multi-scales covering 146 crops in 162 countries, based on a framework coupling GWF assessment with an advanced dynamic decomposition analysis (DDA). Results show that global crop GWF has increased by 1224.1 % in 1961-2018, presenting three stages of GWF growth: an accelerated period of 1961-1980 (P1), a linear period of 1980-1999 (P2), and a decelerated period of 1999-2018 (P3). The spatial and crop-wise patterns of GWF vary for the stages. Crop GWF generally shifts from developed (North America and Europe) to developing (Asia and South America) regions during P1-P2, with crop-wise GWF showing increased proportions in cash crops (Fruits, Vegetables, Oil crops, etc.) for P2-P3 and decreased in food crops (mainly Cereals). China, America, and India exhibit top three national crop GWF amounts and increases (in total 60.3 % of world's GWF growth). Overall, production efficiency and population scale are determinants of growing global crop GWF, with different drivers recognized for the stages. Fertilization level principally drives up world's crop GWF for P1 (50.7 %) while offsets GWF growth in P3 (-8.61 %). Proportion of rural population becomes the main contributor offsetting crop GWF growth especially in P3. Driving pattern of crop GWF evolution varies among continents, countries, and crop categories, as well. Based on these, we recommend optimizing regional/national crop activities expecting to enhance sustainability of global agriculture.

Abstract. Many rivers in the East Asian Monsoon region originates from the Qinghai-Tibet Plateau (QTP), which provide huge amount of fresh water resources for downstream counties. As a region characterized by high altitude and cold weather, distributed hydrological modelling provide valuable knowledge about water cycle and cryosphere of the QTP. However, the lack of streamflow data restricts the application of hydrological models in this data-sparse region. Previous studies have demonstrated the possibility of using remote sensing evapotranspiration (RS-ET) data to improve modelling. However, in the QTP, the mechanisms driving such improvements haven’t been understood thoroughly. In this study, such driving mechanisms were explored through the rainfall-runoff modelling of the Soil and Water Assessment Tool (SWAT) in the Yalong River Basin of the QTP. Three experiments of model calibrations were conducted using streamflow data at the basin outlet, basins averaged RS-ET data of the Global Land Evaporation Amsterdam Model (GLEAM), and the combination of the both data, under the framework of the Generalized Likelihood Uncertainty Analysis (GLUE). The results show that compared with calibration using streamflow data solely, the Nash-Sutcliffe Efficiency of simulated streamflow at 50% quantiles for the calibration using both of streamflow and RS-ET data increased from 0.71 to 0.81 in the calibration period, while in the validation period improved from 0.75 to 0.84, and more observations are embraced by the uncertainty bands. Similar improvements are also found for the ET estimates. Comparison of parameter posterior distributions among the three experiments demonstrated that calibration using both types of observations could increase the number of parameters that posterior distributions are different from assumed uniform prior distribution, indicating the degree of equifinality was reduced. A more comprehensive parameter sensitivity analysis by the Sobol' method were also conducted for reasoning the differences among the three calibrations. Although the number of the detected sensitive parameters are almost same, the sensitive parameter detected based on both types of observations covers surface runoff generation, snow-melting, soil water movement and evaporation processes, while using single type of observations, the identified sensitive parameters are only the ones related the hydrological processed quantified by the observations. From the aspects of model performance and parameter sensitivity, it is demonstrated that not only the model output performs better, but also the characteristics of water cycle are captured more effectively, highlighting the necessity of incorporating RS-ET data for hydrological model calibration in the QTP. Moreover, adopting observations or information about soil property or snow-melting processes to make more reasonable estimates of parameter distribution could further reduce simulation uncertainty under the calibration strategies proposed in this study.

River water surface extent can be extracted from optical and radar satellite images; this is useful for estimating streamflow from space. The radiation characteristics of open water from the visible and microwave bands are different and provide independent information. In this study, for the purpose of improving streamflow estimation from space for data-sparse regions, a method that combines satellite optical and radar images data for streamflow estimation using a machine learning technique was proposed. The method was demonstratedthrough a case study in the river segment upstream of the Ganzi gauging station on the Yalong River, China. Utilizing the support vector regression (SVR) model, the feasibility of different combinations of water surface area derived from Sentinel-1 synthetic aperture radar images (AREA_SAR), modified normalized difference water index derived from Landsat 8 images (MNDWI), and reflectance ratios between NIR and SWIR channels derived from MODIS images (RNIR/RSWIR) for streamflow estimation were evaluated through three experiments. In Experiment I, three models using AREA_SAR (Model 1), MNDWI (Model 2), and a combination of AREA_SAR and MNDWI (Model 3) were built; the mean relative error (MRE) and mean absolute error (MAE) of streamflow estimates corresponding to the SVR model using both AREA_SAR and MNDWI (Model 3) were 0.19 and 31.6 m3/s for the testing dataset, respectively, and were lower than two models using AREA_SAR (Model 1) or MNDWI (Model 2) solely as inputs. In Experiment II, three models with AREA_SAR (Model 4), RNIR/RSWIR (Model 5), and a combination of AREA_SAR and RNIR/RSWIR (Model 6) as inputs were developed; the MRE and MAE for the model using AREA_SAR and RNIR/RSWIR (Model 6) were 0.25 and 56.5 m3/s, respectively, which outperformed the two models treating AREA_SAR (Model 4) or MNDWI (Model 5) as single types of inputs. In Experiment III, three models using AREA_SAR (Model 7), MNDWI, and RNIR/RSWIR (Model 8) and the combination of AREA_SAR, MNDWI and RNIR/RSWIR (Model 9) were built; combining all three types of satellite observations (Model 9) exhibited the highest accuracy, for which the MRE and MAE were 0.18 and 18.4 m3/s, respectively. The results of all three experiments demonstrated that integrating optical and microwave observations could improve the accuracy of streamflow estimates using a data-driven model; the proposed method has great potential for near-real-time estimations of flood magnitude or to reconstruct past variations in streamflow using historical satellite images in data-sparse regions.

Increasingly intensive crop farming has been aggravating global water pollution. It is significant to retrospectively investigate world’s crop production induced water pollution stress over a long time period, facilitating sustainable agriculture in the future. In this study, nearly 60-yr gray water footprint (GWF) dynamics of global crop farming are examined at multi-scales covering 146 crops in 162 countries, based on a long-term reconstruction of gray virtual water content and an advanced dynamic decomposition analysis (DDA). Results show that global crop GWF has increased by 1224.1% in 1961-2018, presenting three stages of GWF growth: an accelerated period of 1961-1980 (P1), a linear period of 1980-1999 (P2), and a decelerated period of 1999-2018 (P3). The evolution pattern of crop GWF varies at spatial and crop-wise scales. A large shift of crop GWF is shown from developed (North America and Europe) to developing (Asia and South America) regions during P1-P2, with crop-wise GWF percentages showing increases in cash crops (Fruits, Vegetables, Oil crops, etc.) for P2-P3 and declinations in food crops (mainly Cereals). China, America, and India exhibit top three national crop GWF amounts and increases (accounting for 60.3% of the world). Globally, production efficiency and population scale are determinants of crop GWF growth for 1961-2018, with contribution proportions of 32.1% and 21.5%, respectively. Fertilization level drives up world’s crop GWF for P1 (50.7%) while contributes to offsetting the growing GWF in P3 (-8.61%). Proportion of rural population becomes a principal driver offsetting crop GWF growth during P1-P3. Driving pattern of crop GWF evolution varies among continents, countries, and different crop categories, as well. Based on these, we expect to provide useful information for enhancing sustainability of global agriculture.

Lakes in arid/semiarid regions face problems of insufficient inflow and degradation of water quality, which threaten the health of the lake ecosystem. Baiyangdian Lake (BYDL), the largest lake in the North China Plain, is confronted with such challenges. The objective of this study was to improve understanding of how changes in water level influence water quality in the BYDL at different temporal scales, especially related to implementations of intermittent environmental water allocation activities in the past two decades, by using data on monthly lake water level, climate factors of precipitation and temperature, and lake water quality. The Mann-Kendall method and continuous wavelet analysis revealed that the lake water level shows a significant decreasing trend after 1967, and the period of 16-year was identified as the principal period for 1950-2018. Based on cross wavelet transform and wavelet coherence analysis, the periodic agreement and coherence between water level and climatic factors decreased after 1997, when environmental water allocations started, indicating that the influences of climatic factors, i.e., precipitation and temperature, became weak. By utilizing the cross-wavelet transform and wavelet coherence analysis methods, the relationships between lake water level and water quality parameters of chemical oxygen demand, ammonia nitrogen, total nitrogen, and total phosphorus were investigated. We found that the change in source and amount of environmental water allocation is one possible reason for the temporal evolution in joint variability between lake water level and water quality. Meanwhile, a dilution effect of freshwater allocated to BYDL was detected in the time-frequency domain. However, the result also indicates that the driving mechanism of water quality is complex due to the combined impacts of water allocation, nonpoint source pollution in the rainy season, and nutrient release from lake sediment. Our findings improve the general understanding of changes in water level in lakes located in arid and semiarid regions under climate change and intensive human activities, and also provide valuable knowledge for decision making in aquatic ecosystem restoration of BYDL and other similar lakes.

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.

In watershed management, it is of great importance to evaluate the risks of nonpoint source (NPS) pollution. In this study, the Nonpoint Source Pollution Risk Index (NSPRI), a multi-factor NPS risk assessment model that was based on the source-sink landscape theory, was proposed and applied in Muzhuhe River Basin, Shandong, China to (1) highlight spatial and temporal variations in the risks from nitrogen and phosphorus losses, and (2) identify how the basin characteristics influenced the risk of nutrient loss. According to the analysis on land use change, the study area is featured with high proportions of forest and agricultural land uses; the area of urban and industrial land had increased considerably from 2000 and 2018. Based on the division of the calculated risk indices on subbasin scale, the area with extremely high risks has decreased from 56,442 ha to 43,922 ha. The average and coefficient of variation (CV) values of NSPRI in the river basin have dropped from 1.3 to 1.1, and from 78.2% to 48.9%, respectively. The distribution of NSPRI suggested an increase in spatial clustering and improvements in the ecological balance. Correlation analysis of the Soil and Water Assessment Tool (SWAT) model (R-2 > 0.68, ENS > 0.59) and NSPRI indicated the applicability of the method used (r > 0.84, p < 0.01). Analysis on the impact of metrics of land use composition, landscape, and environmental settings on NSPRI indicated that the water quality was more significantly correlated with land use composition, landscape pattern and vegetation cover than with flow path distance, soil erodibility, and rainfall erosivity. Moreover, results of redundancy analysis revealed that nutrient loss risk was better explained by land use compositions than by landscape configuration. The assessment method provided scientific support for NPS pollution control from the perspective of source-sink landscape theory.