Abstract: Based on the vegetation photosynthesis model (VPM), we estimated gross primary productivity (GPP) in Gansu Province during 2005-2017, and explored the temporal and spatial variations, variation trend and influencing factors of GPP in different vegetation zones. The results showed that: (1) GPP increased significantly during 2005-2017, with a rate of 6.53 g C·m^-2·a^-1. There were significant differences of GPP among different vegetation zones, with the highest GPP of 1056.03 g C·m^-2·a^-1 in subtropical evergreen broad-leaved forest and the lowest GPP of 164.4 g C·m^-2·a^-1 in temperate desert. (2) In terms of spatial distribution, GPP showed adecreasing trend from southeast to northwest. Qilian Mountains, Qingyang, and southern Gansu were areas with high value. The areas with GPP greater than 500 g C·m^-2·a^-1 accounted for 40.15% of the total area, while the areas with sustained growth of GPP accounted for 83.65%. The areas with poor GPP stability accounted for 30.41%, which mainly distributed in Jiuquan, Wuwei, Lanzhou and Baiyin. The proportion of GPP high aggregation type increased from 29.53% to 30.14%. The high value of GPP increased and became more concentrated in Longnan and Qingyang. (3) The area of woodland increased significantly in Qilian Mountains, Baiyin and Longnan. The area of grassland increased significantly in Jiuquan and Zhangye, while bare land area decreased evidently. The increase of grassland and woodland areas and the reduction of bare land were the main reasons for the increases of GPP in Gansu. (4) The GPP in central Gansu increased with increasing air temperature and precipitation. Rainfall had a positive effect on the increases of GPP in northern Gansu, while temperature had a negative effect on GPP. The change of GPP in southern Gansu showed a contrary pattern.

Key words: gross primary productivity, Gansu Province, VPM model, tempo-spatial change.