Remote Sensing and in-Situ Data for Urban Heat Island Characterization in Oran City

This study addresses the Urban Heat Island (UHI) characterization of Oran agglomeration with the aim of modeling the link between the Land Surface Temperature (LST) previously restituted from the analysis of Landsat satellite images and air temperatures. The results obtained show that the vegetation cover, characterized by NDVI, has a significant influence on the LST, which is considered as a reliable indicator of the UHI. The warmest areas are usually found in dense intra-urban areas and industrial zones. The exploitation of the free data mining software Rattle (graphical interface for data mining using the R software), and the artificial neural networks with multiple linear regression, led to modeling the relationship between surface temperature and air temperature.