Directional Climate Trend, Intensified Intraannual Variability, and Changes in Land Cover Drive the Dynamics of Vegetation Greenness in Peri-Urban China During 2001-2015

摘要

Spatiotemporal dynamics of remote-sensed vegetation indices and derived variables such as land surface phenology have often been studied as a function of climate with implied assumption of negligible land-cover change. However, this assumption is not valid for areas with intensive human activities such as fast-developing countries. To address the impact of land cover on vegetation greenness, we analyzed the trend of EVI_(59), enhanced vegetation index in May-September, within the 25-km~2 patches centered at 674 meteorological stations located mostly in peri-urban areas of China, and the impacts of land cover and intraannual climate variability on the EVI, during 2001-2015 by means of linear mixed-effect model. Impacts of land cover were assessed with the enhancement of sum of the squared difference with respect to the climate model. The climate models explained on average 60% sum of the squared difference of the full models (climate plus land cover), and this proportion reached 83% and 94% for the temperate grassland and the high-cold Tibet, respectively. Including land cover enhanced on average 40% of the sum of the squared difference, and the enhancement is over 60% for the dense-populated warm-temperatedeciduous forest and subtropical-evergreen forest zones. The impact of land cover not only depends on the intensity but also on the type of land-cover change. Forest regrowth in east China and vegetation growth from bare land in temperate desert significantly enhanced greenness but the shift from agriculture to shrubs in temperate grassland may not significantly alter the greenness. We showed that climate variability is important for EVI in all zones except Tibet, and the climate variability contributed on average 64% of climate impacts.