Characterizing Ecosystem Variability Using the Onset of Green-Up Derived from Time-Series AVHRR NDVI Data

摘要

This study explores the use of Vegetation Phenology Metrics (VPMs) derived from satellite remotely sensed measurement to characterize ecosystem variability for the northern China steppes. An Advanced Very High Resolution Radiometer (AVHRR) maximum Normalized Difference Vegetation Index (NDVI) 15-day composite dataset, from 1982 to 1990, was used to estimate the onset date of vegetation green-up for these steppes. The results show a strong linkage between onset date of green-up and climatic factors within the study area. An Ecosystem Variability Index (EVI) was derived using the coefficient of variation of onset date of green-up. The EVI revealed the meadow steppe and forest ecosystems to be the least ecologically variable and the desert steppe to be the most variable. Large areas of typical steppe, near the desert steppe and typical steppe ecotone, exhibited similar EVI values to those of the desert steppe, suggesting greater ecosystem variability in this area of typical steppe. These findings suggest that large expanses of typical steppe in this region might be undergoing ecosystem changes that could result in lower grassland productivity and eventual changes in land use practices. The use of the EVI provides an objective, spatially explicit, and cost-effective means of assessing ecosystem variability and response to varying climatic conditions. This index was used to identify areas within the steppe ecosystem where transition from one stable state to another was most likely to occur, or was already happening.