Quantifying influences of natural factors on vegetation NDVI changes based on geographical detector in Sichuan, western China

Abstract

Abstract: Many studies have revealed that Normalized Difference Vegetation Index (NDVI) is of vital importance for research of ground surface processes and climatic changes. However, identification of the causes responsible for vegetation changes remain challenging. Using Geographical Detector, a new spatial statistical approach, individual and interactive influences of natural factors on vegetation NDVI changes were quantified for the Sichuan province. The optimal characteristics of key natural factors that are beneficial to vegetation growth were determined. The vegetation cover in Sichuan for 2000 and 2015 is good overall; areas with upper-intermediate and high vegetation cover account for more than 88% of the total area. Regions with an NDVI higher than 0.4 are noticeably transformed; regions with upper-intermediate vegetation cover (NDVI from 0.4 to 0.8) tend to decrease significantly, whereas regions with high vegetation cover (NDVI>0.8) tend to increase significantly. Spatiotemporal variation in vegetation cover is very significant. Regions with high vegetation cover are distributed in the northeast part of the Sichuan Basin and on the northwestern Sichuan Plateau, whereas regions with low vegetation cover are distributed in urban agglomerative regions in the central part of the Sichuan Basin. Soil types, elevation, and annual mean temperature can satisfactorily account for vegetation changes. Natural factors have an interactive influence on NDVI. The synergistic effect of natural factors is manifested as mutual enhancement and nonlinear enhancement, and the interaction of two natural factors strengthens the influence of each individual natural factor. Natural factors within a favourable value range or favourable landform factors serve to promote and intervene in vegetation changes, adapt to climatic changes, and buffer against the effects of vegetation changes.