Water quality degradation and eutrophication of lakes are global ecological and environmental concerns, especially shallow lakes. This study collected hydrochemical data from 2935 samples of the Chinese part of Xingkai (Khanka) Lake, based on 40 published papers spanning the period from 2001 to 2023. Using the water quality index (WQI), improved geo-accumulation index (Igeo), and redundancy analysis (RDA), we analyzed the overall contamination characteristics of the water environment in Xingkai Lake. Additionally, we explored the impact of climate change and human activities on the lake’s water quality. The results showed that the annual WQI for Xingkai Lake ranged from 47.3 to 72, with a general downward trend, indicating improving water quality. Notably, the average WQI in May and total nitrogen (TN) content decreased significantly, signaling further improvement in water quality. The average concentration of TN in sediments was 1401.3 mg/kg, reflecting mild contamination. The Igeo values for the heavy metals Hg and Cr were greater than 1, indicating moderate contamination, while the Igeo values for Cd and Pb were between 0 and 1, which is in the range of uncontaminated to moderately contaminated. Land use and climate change (average annual temperature and annual precipitation) were key factors influencing water quality, with cumulative explanatory ratios of 67.3% and 50.1%. This study utilized land-use change as a metric for human activities, highlighting the potential impacts of climate change and human activities on the water quality of Xingkai Lake. It offers vital insights for the sustainable management of Xingkai Lake and provides valuable references into the management of similar transboundary lakes.

Estimating the impacts of climate change on streamflow in the Xiaoxingkai Lake Basin is vital for ensuring sustainable water resource management and transboundary cooperation across the entire Xingkai Lake Basin, a transboundary lake system shared between China and Russia. In this study, 11 Global Climate Models (GCMs) from the Coupled Model Intercomparison Project Phase 6 (CMIP6) under two Shared Socioeconomic Pathways (SSP245 and SSP585) were used to drive the Soil and Water Assessment Tool Plus (SWAT+) model. Streamflow projections were made for two future periods: the 2040s (2021–2060) and the 2080s (2061–2100). To correct for systematic biases in the GCM outputs, we applied the Delta Change method, which significantly reduced root mean square error (RMSE) in both precipitation and temperature by 3–35%, thereby improving the accuracy of SWAT+ simulations. To better capture inter-model variability and enhance the robustness of streamflow projections, we used the Bayesian Model Averaging (BMA) technique to generate a weighted ensemble, which outperformed the simple arithmetic mean by reducing uncertainty across models. Our results indicated that under SSP245, greater increases were projected in annual streamflow as well as in wet and normal-flow seasons (e.g., streamflow in normal-flow season in the 2080s increased by 13.0% under SSP245, compared to 7.0% under SSP585). However, SSP585 produced a much larger relative amplification in the dry season, with percentage changes relative to the historical baseline reaching up to +171.7% in the 2080s, although the corresponding absolute increases remained limited due to the low baseline flow. These findings quantify climate-driven hydrological changes in a cool temperate lake basin by integrating climate projections, hydrological modeling, and ensemble techniques, and highlight their implications for understanding hydrological sustainability under future climate scenarios, providing a critical scientific foundation for developing adaptive, cross-border water management strategies, and for further studies on water resource resilience in transboundary basins.

With the rapid development of the social economy, human activities have increasingly disrupted water environments, and the continuous input of pollutants poses significant challenges for water environment management. Taking the Xiaoxingkai Lake basin as the study area, this paper develops a social–economic–water environment model based on the system dynamics methodology, incorporating subsystems for population, agriculture, and water pollution. The model focuses on four key indicators of pollution severity, namely, total nitrogen (TN), total phosphorus (TP), chemical oxygen demand (COD), and ammonia nitrogen (NH₃-N), and simulates the changes in pollutant loads entering the river under five different scenarios from 2020 to 2030. The results show that agricultural non-point sources are the primary contributors to TN (79.5%) and TP (73.7%), while COD primarily originates from domestic sources (64.2%). NH₃-N is mainly influenced by urban domestic activities (44.7%) and agricultural cultivation (41.2%). Under the status quo development scenario, pollutant loads continue to rise, with more pronounced increases under the economic development scenario, thus posing significant sustainability risks. The pollution control enhancement scenario is most effective in controlling pollutants, but it does not promote socio-economic development and has high implementation costs, failing to achieve coordinated socio-economic and environmental development in the region. The dual-reinforcement scenario and moderate-reinforcement scenario achieve a balance between pollution control and economic development, with the moderate-reinforcement scenario being more suitable for long-term regional development. The findings can provide a scientific basis for water resource management and planning in the Xiaoxingkai Lake basin.