Trend analysis and interactions between surface temperature and vegetation condition: divergent responses across vegetation types

Abstract

Land surface temperature (LST) trends, influenced by climate change, affect vegetation health and productivity, while vegetation, in turn, alters LST by regulating the surface energy balance. These interactions vary by region and vegetation type. In this study, we aimed to (1) examine long-term trends in vegetation conditions and LST over time, and (2) investigate the interactions between vegetation conditions and LST within distinct vegetation types across the Arasbaran Biosphere Reserve. Sentinel-2 spectral-temporal features and the Random Forest model were employed to classify different vegetation types. Time series data for the normalized difference vegetation index (NDVI), normalized difference water index (NDWI), and LST were generated using harmonized Landsat data from 1987 to 2023. Various spatial statistical analyses were applied to address the research questions. The results revealed significant spatial and temporal variations in NDVI, NDWI, and LST among vegetation types. The highest volatility in vegetation conditions occurred in dense and sparse forests, while grasslands exhibited the lowest levels of variability. This variability coincided with an overall increasing trend in NDVI, NDWI, and LST, which was most pronounced in dense forests. Furthermore, a strong negative correlation between NDVI, NDWI, and LST was observed, particularly in croplands. These findings collectively indicate a greening trend in the study area, with forests showing the most pronounced increases. The results also underscore the role of forests and dense vegetation in mitigating projected temperature increases. These insights can inform local land management strategies and decision-making.