Understanding the driving forces for alpine vegetation variations at different permafrost degrading stages is important when the Tibetan Plateau is experiencing climate warming. We applied the modified Frost Number model to simulate frozen ground distributions in the Tibetan Plateau and calculated the maximum thawing depth by the Stefan approach. We classified the simulated frozen ground into three subzones: seasonal frozen ground zone, changing zone, and permafrost zone. We evaluated the effects of precipitation, air temperature, and maximum thawing depth on Normalized Difference Vegetation Index (NDVI) in the subzones across five different stages from 1982 to 2012. The results show that permafrost zone, changing zone, and seasonal frozen ground zone account for about 30.6 percent, 23.3 percent, and 46.1 percent of the study area, respectively. Over the five stages, permafrost areas decreased at fast, slow, fastest, and then slowest rate from stage1 to stage 5, and the large continuous permafrost area has been degraded into pieces. Precipitation is strongly correlated with NDVI and contributes most `to the changes of NDVI. Maximum thawing depth and particularly air temperature show a much smaller correlation and contribute less to the variation rate of NDVI. The findings will have broad applications in investigating the impact of climate and environment changes on alpine vegetation variations in the Tibetan Plateau.

Large-scale engineering projects such as mines and dams cause ecological damage that can persist after construction is complete. The Qinghai–Tibet Railway (QTR) from Golmud to Lhasa was constructed entirely during the Landsat era and began operating in 2006. Therefore, it presents an opportunity for determining both the extent of ecological damage and the time required for some level of natural restoration after a major construction project. We have studied the effect of the QTR construction on vegetation abundance measured using multi-endmember spectral mixture analysis (MESMA) of a time series of Landsat TM/ETM+ images (2001, 2007, 2010) covering the 231 km stretch of the railway from south of Kaixinling Station to north of Chumaerhe Station. We found that the effect of QTR construction on vegetation abundance was limited to within 5.0 km of the tracks, the largest decrease, 2.9%, occurring within 0.125 km of the tracks from 2001 to 2007. There was only 0.4% further decrease in vegetation abundance from 2007 to 2010. We attributed the decrease in vegetation abundance within 0.125 km of the tracks mostly to the accumulation of drifting sand resulting from the barrier of the railway and the measures adopted to keep the sand away from the railway under the prevailing west or northwest winds. It appears that in this sensitive cold desert, vegetation damage was limited to the period of railway construction.

The normalized difference vegetation index (NDVI) can depict the status of vegetation growth and coverage in grasslands, whereas coarse spatial resolution, cloud cover, and vegetation phenology limit its applicability in fine-scale research, especially in areas covering various vegetation or in fragmented landscapes. In this study, a methodology was developed for obtaining the 30 m annual maximum NDVI to overcome these shortcomings. First, the Landsat NDVI was simulated by fusing Landsat and MODIS NDVI by using the enhanced spatial and temporal adaptive reflectance fusion model (ESTARFM), and then a single-peaked symmetric logistic model was employed to fit the Landsat NDVI data and derive the maximum NDVI in a year. The annual maximum NDVI was then used as a season-independent substitute to monitor grassland variation from 2001 to 2022 in a typical area covering the major vegetation types in the Qinghai Lake Basin. The major conclusions are as follows: (1) Our method for reconstructing the NDVI time series yielded higher accuracy than the existing dataset. The root mean square error (RMSE) for 91.8% of the pixels was less than 0.1. (2) The annual maximum NDVI from 2001 to 2022 exhibited spatial distribution characteristics, with higher values in the northern and southern regions and lower values in the central area. In addition, the earlier vegetation growth maximum dates were related to the vegetation type and accompanied by higher NDVI maxima in the study area. (3) The overall interannual variation showed a slight increasing trend from 2001 to 2022, and the degraded area was characterized as patches and was dominated by Alpine kobresia spp., Forb Meadow, whose change resulted from a combination of permafrost degradation, overgrazing, and rodent infestation and should be given more attention in the Qinghai Lake Basin.

The simulation and prediction of coastline changes are of great significance for the development and scientific management of coastal zones. Coastline changes are difficult to capture completely but appear significantly periodic over a long time series. In this paper, the transformer model is used to learn the changing trend of the coastline so as to deduce the position of the coastline in the coming year. First, we use the distance regularization level set evolution (DRLSE) model for instantaneous waterline extraction (IWE) from preprocessed Landsat time-series images from 2010–2020 in Weitou Bay, China. Then, tidal correction (TC) is performed on the extracted instantaneous waterline dataset to obtain coastlines projected to a single reference tidal datum. Finally, the coastline datasets from 2010–2019 are used for model training, and the coastline in 2020 is used for accuracy assessment. Three precision evaluation methods, including receiver operating characteristic curve matching, the mean offset, and the root mean square error, were used to verify the predicted coastline data. The receiver operating characteristic curve was specifically designed and improved to evaluate the accuracy of the obtained coastline. Compared with the support vector regression (SVR) and long–short-term memory (LSTM) methods, the results showed that the coastline predicted by the transformer model was the closest to the accurate extracted coastline. The accuracies of the correct values corresponding to SVR, LSTM, and transformer models were 88.27%, 94.08%, and 98.80%, respectively, which indicated the accuracy of the coastline extraction results. Additionally, the mean offset and root mean square error were 0.32 pixels and 0.57 pixels, respectively. In addition, the experimental results showed that tidal correction is important for coastline prediction. Moreover, through field investigations of coastlines, the predicted results obtained for natural coastlines were more accurate, while the predicted results were relatively poor for some artificial coastlines that were intensely influenced by human activities. This study shows that the transformer model can provide natural coastline changes for coastal management.

Understanding the vegetation phenology and its variations in the Tibetan Plateau is critical to the study of ecological responses to global climate change. In this study, several pre-processed methods or techniques were applied to filter the Global Inventory Modelling and Mapping Study’s Normalized Difference Vegetation Index (GIMMS NDVI) data from 1982 to 2006, and construct the daily NDVI series. Firstly, vegetation and non-vegetation were determined by NDVI quantity contour, and cloud-covered pixels were also eliminated by NDVI change characteristics in a year. Then, the NDVI series were filtered by three-standard deviation and Savitzky-Golay method. Finally, the Savitzky–Golay method was employed to fit and construct the daily NDVI series. These methods guarantee a more reliable subsequent calculation of subsequent vegetation phenology. The vegetation phenology parameters including the start of growth season (SOG), the end of growth season (EOG), the lengths of growth season (LOG) and the absolute increase in vegetation (AIV), defined as the difference between the maximum NDVI and the NDVI for SOG in a year, were derived from the daily NDVI series based on the maximum ratio threshold method and their variations were analysed. The results showed that the SOGs were gradually delayed from the southeast to the northwest of the Tibetan Plateau, but the distribution pattern of the EOGs was opposite to that of the SOGs. From 1982 to 2006, SOGs were advanced approximately 3–18 days and EOGs delayed around 0–24 days in the southeast, whereas AIVs decreased around 0–0.3. In the northwest, these phenology parameters followed inverse trends compared with those of the southeast. Over the 25-year period, LOG changes had no constructive or active effects on the vegetation absolute increase. These complex phenological shifts were mainly due to the spatial differences in the environmental changes. However, in some extent, they might be related to the vegetation itself, such as its fractional cover. These findings may help to understand the alpine vegetation responds to climate change in the Tibetan Plateau.