Lawrence Berkeley National Laboratory

Terrestrial gross primary productivity (GPP) is an important flux that drives the global carbon cycle. However, quantifying the trend and the control factor of GPP from the pixel level to the regional level is still a challenge. We generated monthly GPP dataset using the vegetation photosynthesis model and calculated the interannual linear trend for China during 2000-2016. The Breaks For Additive Seasonal and Trend method was applied to detect the timing of breakpoint and trends shift of monthly GPP, while boosted regression tree analysis was used to identify the most important factor and its relative influence on GPP based on gridded leaf area index (LAI), aerosol optical thickness, and NCEP-DOE Reanalysis II meteorological data. The results show that annual mean GPP was significantly (P < 0.001, R 2 = 0.78) increased, especially in the Loess Plateau and South China, from 2000 to 2016. The change rate of annual mean GPP declined from 18.82 g C m-2 yr-1 in 2000-2008 to 3.48 g C m-2 yr-1 in 2008-2016. About 55.4% of the breakpoints occur between 2009 and 2011 and was mainly distributed in Qinghai-Tibet Plateau, Central China, Southwestern China, and South China, and negative oriented GPP trends variation type still accounts for about 28.76%. LAI and temperature related factors generally had the highest relative influence on GPP in the north part and south part of China, respectively. Our study indicates that the ecological restoration projects and rapid urbanization have respectively induced the most obvious increase and decrease trends of GPP in China. Land cover change and climate change are the main reasons for GPP dynamics in the north and south part of China, respectively.