[Spatio-temporal Variation in Net Primary Productivity of Different Vegetation Types and Its Influencing Factors Exploration in Southwest China]

Spatiotemporal variations and its driving factors of NDVI in Northwest China during 2000-2021

Northwest China (WTL) is an essential ecological barrier zone of China, an important node of the "Silk Road Economic Belt," and a crucial bearing area for China's execution of the "One Road and One Belt" and "Going Global" strategies. However, its ecology is exceedingly fragile and particularly vulnerable to climate change and human interference. This study explored the spatiotemporal evolution characteristics of vegetation in WTL using NDVI data and investigated its drive mechanisms by geodetector, partial correlation analysis, and residual trend analysis methods. As well as forecasting the trend for vegetation changes. The findings demonstrated that (1) the change in NDVI manifested an overall improvement trend and the distribution in space of NDVI rose from the center to the periphery. 57.07% of the area had a sparse cover of vegetation (NDVI between 0 and 0.2). In addition, about 49% of regions had deterioration tendencies, which were mainly aggregated in HX, QCXDB, QCXDN, and the eastern of QCXQN and QCXXB. (2) The NDVI's shifting trend was unsustainability, and the region of uncertain future accounted for 57.45% of the total, with apparent unsustainability features. (3) The key parameters influencing NDVI spatial distribution were Pre (precipitation), vegetation type, land use type, and soil type. The interaction between two factors enhanced the influence of any single element, which appeared as bivariate and nonlinear enhancements. (4) Both climate variations and human activities have been recognized as key variables affecting NDVI growth. NDVI variance in 73.02% of areas was influenced by the combined effects of climate variations and human activities. However, human activities were the most influential element in NDVI growth, with the relative contributions of 80.28% (19.72% of which was caused by climate variations). These results can be conducive to deepening insights into the local vegetation status, identifying the mechanisms driving vegetation change, and providing scientific recommendations for WTL's ecosystem restoration measures based on actual situations.