Climate and Aridity Measures Relationships with Spectral Vegetation Indices across Desert Fringe Shrublands in the South-Eastern Mediterranean Basin

Abstract

Abstract Mediterranean regions are hot spots of climate change, where the expected decrease in water resources threatens the sustainability of shrublands at their arid margins. Studying spectral vegetation indices relationships with rainfall and Potential Evapotranspiration (PET) changes across Mediterranean to arid transition zones is instrumental for developing methods for mapping and monitoring the effects of climate change on desert fringe shrublands. Here we examined relationships between 17 spectral vegetation indices (VIs) and four climate and aridity measures: rainfall, PET, Aridity Index (AI) and Water Deficit (WD) calculated at accumulation lags between 1 and 6 months. For this purpose, VIs for 38 sites (100x100 meters each) representing less disturbed areas were extracted from Sentinel 2A images for 3 years with high (2016), low (2017), and average (2018) annual rainfall. Most of the VIs had shown the highest correlation with the four climate and aridity measures at 2 months accumulation interval. While NDVI relationships with climate measures gained the widest use, our data suggest that indices combining NIR and SWIR bands better correlate climate parameters. AI is one of the leading annual measures of dryness worldwide, when calculating it monthly, WD was found to better represent the balance between precipitation and PET across the climate transition zone, and to be better correlated with VI’s. Relationships between NIR and SWIR VIs and Water Deficit may thus facilitate improvements in monitoring and mapping desert fringe shrublands responses to climate change if supported by similar results from wider areas.