Abstract: Vegetation phenology is the most significant and intuitive indicator of climate change, and it is of great significance for understanding the interaction between climate change and vegetation phenology. The response and feedback of different ecosystems to climate change are different. It is of great significance for understanding the mechanism of phenological change to explore the relationship between spatiotemporal variation of vegetation phenology and its driving factors in various ecosystems under climate change. In this study, the Bayinbuluke with a variety of ecosystems was selected as the study area. The MODIS NDVI time-series data during the period of 1989-2011 were used to get the spatial distribution and spatiotemporal variation of vegetation phenology, and to analyze its relationship with elevation and climate factors. The results showed that: ① The beginning time of vegetation growth season was mainly from early April to mid-May, and the ending time was mainly from mid-September to mid-October; ② The beginning time of vegetation growth season in most regions of the study area became earlier during the period of 2001-2017, and that of the grassland, meadow and swamp became significantly earlier except for the alpine vegetation. The ending time of vegetation growth season became earlier in the north of the study area but was delayed in the south, and that of the alpine vegetation and grassland became significantly earlier; ③ On the whole, the beginning time of vegetation growth season was delayed but the ending time became earlier with the increase of elevation, and phenological difference among different ecosystems was decreased with the increase of elevation; ④ There was a negative correlation between the beginning time of vegetation growth season and the spring temperature. The beginning time of growth season of alpine vegetation was significantly affected by temperature in May, that of grassland and swamp by temperature in April, and that of meadow by temperature in April and May.

Key words: vegetation phenology, NDVI, spatial pattern, climate change, Bayanbuluk, Xinjiang