Ground rainfall observations are essential for improving gauge-adjusted satellite precipitation estimates, particularly in the Global Precipitation Measurement mission era. However, over 60% of countries lack a complete surface-based rainfall gauge network, severely constraining the accuracy of global satellite precipitation retrievals. Here we present a model that integrates regional-scale intelligent optimization, topographic analysis, and an end-to-end neural network to merge multi-source precipitation data and reconstruct global spatiotemporal precipitation fields with resolved data gaps across scales. Our evaluation demonstrates that the model substantially improves satellite precipitation accuracy, increasing the correlation coefficient from 0.66 to 0.77. In particular, satellite estimates at high rain rates and over complex terrains show marked improvement when the model is applied. These findings suggest that the proposed approach holds strong potential for future applications in storm monitoring and flood forecasting, especially in mountainous regions with sparse observational data.

There are a large number of lakes with beaded distribution in the semi-arid areas of the Inner Mongolian Plateau, and some of them have degraded or even disappeared during the past three decades. We studied the reasons of the disappearance of these lakes by determining the way of replenishment of these lakes and the impact of the natural-social environment of the basin, with the aim of saving these gradually disappearing lakes. Based on remote sensing image and hydrological analysis, this paper studied the recharge of Daihai Lake and Huangqihai Lake. The deep learning method was used to establish the time-series of lake evolution. The same method was combined with the innovative woodland and farmland extraction method to set up the time-series of ground classification composition in the basins. Using relevant survey data, combined with soil water infiltration test, water chemical, and isotopic signature analysis of various water bodies, we found that the Daihai Lake area is the largest in dry season and the smallest in rainy season and the other lake is not satisfied with this phenomenon. In addition, we calculated the specific recharge and consumption of the study basin. These experiments indicated that the exogenous groundwater is recharged directly through the faults at the bottom of Daihai Lake, while the exogenous groundwater is recharged in Huangqihai Lake through rivers indirectly. Large-scale exploitation of groundwater for agricultural irrigation and industrial production is the main cause of lake degradation. Reducing the extraction of groundwater for agricultural irrigation is an important measure to restore lake ecology.

The relationship between groundwater and lakes in Qaidam Basin is often overlooked. Therefore, we employed Landsat satellite images and meteorological data to investigate the causes of lake expansion through model calculation and statistical analysis and then determine groundwater sources through isotope analysis (²H, ³H, and ¹⁸O). In the two study periods of 2003–2011 and 2011–present, temperature, precipitation, and runoff increased at a steady rate, whereas the expansion rate of Tuosu Lake increased from 1.22 km²/yearr to 3.38 km²/yearr. This significant increase in the rate of lake expansion reflects the substantial contribution of groundwater to lake expansion. The groundwater contribution to the lake includes not only the glacial meltwater that infiltrates the piedmont plain but also other, more isotopically deleted water sources from other basins. It is speculated that the 2003 M_s 6.4 earthquake in the northwest of the Delingha region was a possible mechanism for lake expansion. Earthquakes can enhance crustal permeability and keep fractures open, which promotes groundwater contribution to lakes and in turn causes rapid lake expansion and an increased groundwater level. This study is important for understanding the sources, circulation, and evolution of groundwater in Qaidam Basin.

Great inland freshwater lakes play an important role in regulating inland water resources, and the usage of synthetic aperture radar (SAR) images for the accurate waterline mapping is an effective technical means to study the dynamic changes of great inland lakes. In this article, the Danjiangkou (DJK) reservoir is selected as a study case, and a novel waterline mapping method with four main parts is proposed to monitor the water area dynamically. First, a coarse segmentation method is implemented to extract the initial waterline. Second, a strategy of division in local regions is given to speed up the subsequent processes. Third, a combination of a speckle filter and an improved geometric active contour model is used for refined segmentation. Finally, a change detection method is used to study the changing lake. Furthermore, six SAR images obtained by the Gaofen-3 (GF-3) and Sentinel-1A (S-1A) satellites in the DJK reservoir, Hongze lake and Poyang lake are tested to verify the universality of the proposed water area extraction method. The results demonstrate excellent performances with an accuracy of over 97% and an average contour offset under 0.7 pixels. Besides, the time-series analysis of the DJK reservoir is applied based on the mapped waterlines of 37 SAR images collected from January to December in 2017. Comparing with the changing tendency of the water level surveyed in the DJK reservoir, the waterline mapping results and the filed survey data have great consistency, which further proves the validity of the proposed method, also presents the significant potential of the GF-3 and S-1A SAR images for managing the water resource.