Observed daily temperature and precipitation from 121 meteorological stations and satellite-based 8- day average gross primary productivity (GPP) from MOD17A2 are utilized to develop the linear correlation models between 8-day accumulated average temperature, maximum temperature, minimum temperature, precipitation and accumulated GPP in the monsoon zone in Northern China during 2000-2013. Based on the derived thresholds and coefficients of these models, variability in the starting date of GPP accumulation, length of GPP accumulating period, ending date of GPP accumulation as well as GPP accumulation rate on forest, grassland and cropland ecosystems are investigated under two Representative Concentration Pathways (RCP4.5 and RCP8.5) of the Regional Spectral Model (RSM). Finally, the substantial impacts of climate changes including maximum, average, minimum temperature and precipitation on ecosystem productivity are evaluated. Results suggest that average and minimum temperature can predict GPP more accurately than maximum temperature and precipitation. Besides, the starting and ending dates of GPP accumulation are sensitive to the variability in four climatic factors whereas the GPP accumulation period and rate are more sensitive to the variability in average and maximum temperature. Additionally, future climate changes tend to prolong the GPP accumulation period and increase the GPP accumulation rate, thus increasing GPP.