A geographically weighted regression approach to detect divergent changes in the vegetation activity along the elevation gradients over the last 20 years

Abstract The Wuyi Mountains National Nature Reserve, a World Heritage site listed by the United Nations Educational, Scientific and Cultural Organisation (UNESCO), preserves the world’s largest and most representative subtropical forests at the same latitude. However, little is known about the spatio-temporal patterns in vegetation activity along the elevation gradients over the last two decades in response to global warming. In this study, we analysed the spatio-temporal patterns of vegetation greenness and their elevation gradient dynamics during 1998–2018. We found a relatively high vegetation coverage in the study area, with an averaged normalised difference vegetation index (NDVI) between 0.79 and 0.86 and 78.1% of the study area under green cover during this period. We employed geographically weighted regression (GWR) to explore the spatio-temporal patterns of the vegetation activity along the elevation gradients during 1998–2018. We observed that the average elevation for the negative relationships of vegetation elevation was higher than that for the positive relationships between elevation and NDVI. Although our results are not fully consistent with the current homogenisation trend in vegetation activity along the elevation gradients, we found that 35% of the study area showed increased homogenisation. Our findings of highly variable patterns in the vegetation activity of elevation gradients will help mitigate the effects of global warming and aide in biodiversity conservation in the Wuyi Mountains National Nature Reserve and similar areas at the same latitude.