How the characteristics of land cover changes affect vegetation greenness in Guangdong, a rapid urbanization region of China during 2001-2022.

To evaluate the quantitative impacts of land cover change on vegetation greenness in the significantly human-impacted subtropical region, the characteristics of land cover change were explored by land use dynamic degree, transition matrix and normalized entropy. Various methods including Standardized coefficient, LMG (Lindeman-Merenda-Gold), GEN (Genizi measure) and CAR (Correlation-Adjusted Marginal Correlation) were employed to estimate the contributions of land cover changes on vegetation greenness using MODIS data during 2001-2022 in Guangdong. The conclusions revealed that land cover changes exhibited obvious temporal characteristics in Guangdong with a significantly increasing trend of normalized entropy indicating a more balanced distribution of land cover types under human intervention. NDVI (Normalized Difference Vegetation Index) tended to increase likely due to the large-scale increase in evergreen forest. With regard to the contributions of impact factors on vegetation greenness, the contributions evaluated by LMG, GEN and CAR showed that the natural variation of NDVI accounted for the major contribution (> 33%), while the changes of evergreen forest and grassland had the highest contribution (> 37%) according to Standardized coefficient. These differences were mainly due to the characteristics of land cover changes in Guangdong, the correlations among impact factors and the inherent attributions of the methods. Moreover, the expansions of evergreen forest and urban at the expense of the reductions of grassland and cropland also had significant impacts on NDVI (> 10%) according to LMG, GEN and CAR indicating that human-induced land cover changes had remarkable influences on NDVI.