Appropriate plant indicators derived from satellite images to investigate the impact of climatic parameters on forest cover in Northern Zagros, Iran

This study aimed at selecting the most appropriate vegetation indices extracted from satellite images to study the effect of climate change on forest cover in the Sardasht, W. Azerbaijan province, Iran. Therefore, 25 plant indices were initially calculated from Landsat time-series images during 1988-2017, followed by applying principal component analysis (PCA), which led to the selection of nine more effective plant indices. The selected indices included seven indices in the first axis and two in the second axis. These two axes explained ca. 91% of the variance of the studied indicators. In addition, linear correlation coefficients between selected indices and climatic parameters were investigated using the data of the Sardasht synoptic station. The results showed that the indicators of the first and second axes were directly and inversely correlated with spring rainfall and temperature, respectively. The Soil Adjusted Vegetation Index (SAVI) showed 49% correlation with spring rainfall, while the Red Green Ratio Index (RGRI) revealed 51% correlation with the average spring temperature during the 30-year period. The association of SAVI and RGRI indices with precipitation and temperature in the last three decades showed a correlation of 84% and 51%, respectively, mainly due to the matching effects of precipitation and spring temperature in the first decade. However, correlation coefficients of both indices decreased to less than 50 and 10%, respectively, in the second and third decades. According to 1) the characteristics of the study area and the involvement of other factors that affect forest cover as well as 2) the opposite or similar effects of climate parameters on forest cover, application of the mentioned indicators to survey the effect of climate change on the Northern Zagros forest cover showed low to moderate correlations.