Temporal dynamics of satellite-derived vegetation pattern and growth in an arid inland river basin, Tibetan Plateau

Exploring alpine vegetation dynamics and its driving factors help to predict changes in hydrothermal circulations at high elevations in the future. Currently, there is a lack of systematic research on this topic in the permafrost regions at a basin scale. Spatiotemporal variations of vegetation pattern and growth estimated from satellite remote sensing images at a 30 m spatial resolution on the Google Earth Engine cloud platform were investigated in the Shule River headwater region from 2000 to 2021. The results showed shrinking trends in alpine grassland (AG) and wetland (AW) area, while an expanding trend in alpine desert (AD) area. AD moved to lower altitudes and occupied AG and AW area, possibly owing to mountain permafrost degradation and extreme precipitation. Vegetation patches became fragmented and complex, reflected by increases in patch number and shape index. Vegetation growth, as approximated by the normalized difference vegetation index (NDVI), increased significantly in 42 % of the vegetated area (p < 0.05), particularly in river valleys. In addition, NDVI was negatively correlated with solar radiation over 30 % of the vegetated area (p < 0.05). At the regional scale, vegetation NDVI was positively correlated with root-zone soil moisture and grazing intensity (p < 0.05). The findings suggested that AG and AW at their upper limit of distribution were vulnerable to soil erosion, and the inter-annual variation of vegetation growth was affected by soil moisture and grazing. Moreover, future climate warming may cause alpine vegetation decline by increasing active layer thickness in the absence of adequate precipitation supply.