Long-term Global Land Surface Phenology Derived from AVHRR and MODIS Observations and Its response to Climate Changes

Land surface phenology has been widely retrieved from satellite observations at regional and global scales. Since it is an ideal indicator of recent climate changes, spring vegetation greenup has been frequently applied to explore the warming climate impacts in middle-high latitudes. However, we understand poorly the diverse responses of sequent phenological indicators that comprise an entire vegetation growing cycle to climate changes at broad environments. It is hypnotized that the timing of individual phenological indicators in a seasonal cycle may be independently advanced, delayed, or unchanged in responses to climate change. Integrating a sequence of key phenological indicators is expected to more effectively reflect long-term climate variation in various seasons. It is also more effective to track, trace and project the climate impacts as climate change continues. This study detected global land surface phenology from AVHRR and MODIS from 1982-2010. Specifically, based on a dataset of daily enhanced vegetation index (EVI) at a spatial resolution of 0.05 degrees, we simulated the seasonal vegetative trajectory for each individual pixel and then detected the phenological indicators including onset of greenness increase, onset of greenness maximum, onset of greenness decrease, onset of greenness minimum, the onset of middle greenup phase, the onset of middle senescent phase, growing season length, magnitude EVI, and growing season aggression in EVI. Further, we examined the interannual variations and trends of the phenological indicators from 1982-2010. Meanwhile, the phenological variations were directly linked to long-term global precipitation and temperature. The results indicate (1) spring green up is consistently advanced in some regions, such as Alaska; (2) vegetation greenup phase become short in southern hemisphere; (3) interannual variation in vegetation growth is significantly increased during summer and autumn globally; (4) these temporal and spatial patterns effectively reflect the variations in climate variables.