Multivariate Kernels in GWR Model to Identify Climatic Micro-Basins

The aim of this work is to quantify the effect of temperature and precipitation on Normalized Difference Vegetation Index (NDVI). The relations between variables are analysed at the local level by estimating three GWR models. In an innovative way we apply some multivariate kernels to GWR models to better understand and examine the effect of spatial dimensions on spatial weights. Estimated models show a strong spatial variability of the parameters which highlights the importance of local models for this kind of studies.