The use of the global navigation satellite system (GNSS) for monitoring changes in terrestrial water storage (TWS) is growing. However, the density of GNSS stations is sparse in most areas, and the widely used Green’s function (GF) method cannot provide a satisfactory resolution for inversion problems. As the Slepian basis function (SBF) method has been successfully used for gravity inversions, GNSS geodesists have recently applied the SBF method to GNSS displacements. However, the evaluation of TWS differences inferred from GNSS using the SBF and the GF has been rarely assessed. In this study, we use both the GF and the SBF methods to investigate the TWS in Yunnan by using more than ten years (2010–2021) of GNSS observations. We observe a remarkable consistency between the two inversion methods employed for GNSS data, as well as a strong agreement with equivalent water heights (EWH) inferred from the Gravity Recovery and Climate Experiment (GRACE) and GRACE Follow-On (GFO), hydrological model (GLDAS-NOAH), and precipitation data, despite variations in the amplitude. Furthermore, affected by monsoon climate and topography, Yunnan is prone to drought. We analyze the main nine drought events, evaluate the connection between GNSS-inverted TWS during 2010–2021, and find that almost all droughts in Yunnan occurred during the "trough period" within the GNSS-DSI.

The runoff change in a large river is significant for hydrological event monitoring, such as floods and droughts. In this paper, we attempt to invert the river flow change based on the surface displacements observed by the continuous global navigation satellite system (GNSS) stations in the Wuhan section of the Yangtze River (WYR). We establish a two‐dimensional model accounting for deformation from a line load in an elastic half‐space to estimate the annual water load change from the GNSS data. The results show that the inverted line load from the annual maximum displacement change observed by the GNSS is , which is in good agreement with the river load from the hydrological station. This paper demonstrates a low‐cost and practical method to river flow research in local areas, which is also a creative application for the GNSS.

Over the past 18 years, the water level of Qinghai Lake has risen by more than 4 m, resulting in a total mass increase of approximately 22 Gt. This increase may have impacted the surface deformation and fault activity within a 100 km radius of the lake. The deformation induced by water load changes is captured by the Global Navigation Satellite System (GNSS) stations installed around Qinghai Lake. In this study, we develop a three‐dimensional elastic half‐space model to explain more following surface displacement and stress change on faults caused due to lake level rise. Our results highlight that the vertical crustal movement around Qinghai Lake is mainly characterized by uplift. Additionally, we calculated the variation in Coulomb stress change on several faults nearby and found that the largest Coulomb stress change caused by the increase in lake loading is 27.8 kPa. The Coulomb stress change will further increase with the ongoing lake level rise. Assuming the projected water level rise of 14 m (to 3,207 m) by 2,035 under high‐emission scenarios, our model predicts that the resulting hydrological load could amplify Coulomb stress changes on nearby faults by over threefold compared to current values. This escalation may contribute to stress conditions that are associated with increased potential for seismic activity. Furthermore, our analysis suggests that long‐term fluctuations in lake filling events may modulate vertical crustal motion and influence stress conditions on fault planes, potentially contributing to seismic activity patterns.

Most existing Simultaneous Localization and Mapping (SLAM) systems rely on the assumption of static environments to achieve reliable and efficient mapping. However, such methods often suffer from degraded localization accuracy and mapping consistency in dynamic settings, as they lack explicit mechanisms to distinguish between static and dynamic elements. To overcome this limitation, we present BMP-SLAM, a vision-based SLAM approach that integrates semantic segmentation and Bayesian motion estimation to robustly handle dynamic indoor scenes. To enable real-time dynamic object detection, we integrate YOLOv5, a semantic segmentation network that identifies and localizes dynamic regions within the environment, into a dedicated dynamic target detection thread. Simultaneously, the data association Bayesian mobile probability proposed in this paper effectively eliminates dynamic feature points and successfully reduces the impact of dynamic targets in the environment on the SLAM system. To enhance complex indoor robotic navigation, the proposed system integrates semantic keyframe information with dynamic object detection outputs to reconstruct high-fidelity 3D point cloud maps of indoor environments. The evaluation conducted on the TUM RGB-D dataset indicates that the performance of BMP-SLAM is superior to that of ORB-SLAM3, with the trajectory tracking accuracy improved by 96.35%. Comparative evaluations demonstrate that the proposed system achieves superior performance in dynamic environments, exhibiting both lower trajectory drift and enhanced positioning precision relative to state-of-the-art dynamic SLAM methods.