摘要
Vegetation phenology is an efficient bio-indicator of impacts of climate changes and a key parameter for understanding and modelling vegetation-climate interactions and their implications on carbon cycling. This paper analyzes the relation between satellite-based measures of vegetation phenology and monthly temperature according to ecoregions in Tibetan Plateau from 2000 to 2009. Firstly, an algorithm of dynamic threshold is applied to extract start of season (SOS) and end of season (EOS) based on Moderate Resolution Imaging Spectroradiometer (MODIS) 250m 16-days normalized difference vegetation index (NDVI) data set. Secondly, the phenological responses to climate warming were assessed by the linear regression of phenological dates against the current monthly temperature from 2000 to 2009. The vegetation ecosystems are sensitive to climate warming in Tibetan Plateau. For each degree warming in spring,the spring phenology would be advanced by 2.72-9.72 days for different ecoregions. For each degree warming in the fall, the autumn phenology would be delayed by 5.69-6.78 days for different ecoregions. Compared to steppe, meadow shows a larger advance in spring phenology and delay in fall phenology under climate warming condition.