Abstract
This study investigated spatiotemporal patterns of growing season gross primary productivity (GPP(GS)) across three vegetation types in mid-to-high latitude East Asia from 2001-2020.Growing season parameters and GPP were extracted from MODIS satellite data and combined with meteorological data to examine climate-vegetation relationships through trend analysis and correlation methods. GPP(GS) increased significantly overall (4.12 gC/m2/yr), with deciduous broad-leaved forest (DBF) having highest productivity (1035.52 gC/m2), followed by deciduous needle-leaved forest (DNF) (830.83 gC/m2) and grassland (800.62 gC/m2). A critical divergence occurred around 2014, when grassland and DNF growth rates declined substantially while DBF maintained steady increases. Phenological factors showed limited explanatory power for GPP variations, albeit GPP are sensitive to vegetation peak growth time for all three vegetation types. Climate analysis identified relative humidity (RH) as the dominant driver, with the previous year's growing season RH showing around 35.91% stronger positive correlations than current year values across all vegetation types; the difference is highest in DNF and the least in grassland. We conclude that the legacy effects of atmospheric moisture conditions explained the 2014 divergence, highlighting the increasing importance of water availability under global warming. Increases in atmospheric dryness accompanied by temperature increases will affect vegetation carbon storage and the societal-economic services provided by these ecosystems.