In recent decades, landslide displacement forecasting has received increasing attention due to its ability to reduce landslide hazards. To improve the forecast accuracy of landslide displacement, a dynamic forecasting model based on variational mode decomposition (VMD) and a stack long short-term memory network (SLSTM) is proposed. VMD is used to decompose landslide displacement into different displacement subsequences, and the SLSTM network is used to forecast each displacement subsequence. Then, the forecast values of landslide displacement are obtained by reconstructing the forecast values of all displacement subsequences. On the other hand, the SLSTM networks are updated by adding the forecast values into the training set, realizing the dynamic displacement forecasting. The proposed model was verified on the Dashuitian landslide in China. The results show that compared with the two advanced forecasting models, long short-term memory (LSTM) network, and empirical mode decomposition (EMD)–LSTM network, the proposed model has higher forecast accuracy.

We investigated the effect of mass loading (atmospheric, oceanic and hydrological loading (AOH)) on Global Positioning System (GPS) height time series from 30 GPS stations in the Eurasian plate. Wavelet coherence (WTC) was employed to inspect the correlation and the time-variable relative phase between the two signals in the time–frequency domain. The results of the WTC-based semblance analysis indicated that the annual fluctuations in the two signals for most sites are physically related. The phase asynchrony at the annual time scale between GPS heights and AOH displacements indicated that the annual oscillation in GPS heights is due to a combination of mass loading signals and systematic errors (AOH modelling errors, geophysical effects and/or GPS system errors). Moreover, we discuss the impacts of AOH corrections on GPS noise estimation. The results showed that not all sites have an improved velocity uncertainty due to the increased amplitude of noise and/or the decreased spectral index after AOH corrections. Therefore, the posterior mass loading model correction is potentially feasible but not sufficient.

Timely monitoring of urban ecological environments is essential for sustainable development. This study addresses the need to understand the complex internal mechanisms of ecological quality change, especially when indicators present contradictory trends. It evaluates the spatiotemporal evolution of ecological quality in Hangzhou, China, from 2016 to 2024, to identify the specific drivers of its improvement. The study utilized Landsat 8 data processed on the Google Earth Engine platform. Four key indicators were extracted: greenness (NDVI), wetness (WET), dryness (NDBSI), and heat (LST). These indicators were then integrated using Principal Component Analysis (PCA) to construct the comprehensive Remote Sensing Ecological Index (RSEI). The findings reveal that Hangzhou's mean RSEI increased by 8.07%, rising from 0.5249 in 2016 to 0.5710 in 2024. This overall enhancement was marked by a 9.55% expansion of the “Excellent” ecological grade area. Spatially, the city maintained a stable “southwest-high, northeast-low” ecological gradient. This ecological gain presents a significant finding. The improvement was not driven by an increase in greenness. The city-wide mean NDVI remained stable with a 0.15% decrease. Instead, the RSEI increase was caused by the mitigation of human-induced stress. This is evidenced by a sharp 17.55% decrease in dryness (NDBSI) and a 9.77% decrease in heat (LST). Hangzhou’s ecological progress was achieved by alleviating urban pressures, such as the heat island effect and impervious surface impacts, rather than by expanding vegetation. This study validates the RSEI's utility for capturing complex improvements in mature urban ecosystems.