Spatiotemporal analysis of land surface temperature using multi-temporal and multi-sensor image fusion techniques

Abstract Urban Heat Island (UHI) is a major challenge in urban environments that affects human activities. This phenomenon is caused and changed by various factors in urban environments, in which identification of such parameters requires fine-scale satellite images. This study, therefore, aimed at investigating spatiotemporal changes of UHIs as well as identification of important factors using remote sensing image fusion techniques in Rasht. The image dataset comprises of Landsat 5, 7, 8 and MODIS images from 2001 to 2018. After pre-processing, multi-temporal and multi-sensor image fusion techniques were used to retrieve Landsat 7 images as well as Landsat-like medium spatial resolution images. The effects of built-up areas and surface biophysical characteristics such as brightness, greenness and wetness were also examined on LST changes through time. The results showed that multi-temporal and multi-sensor image fusion methods provide appropriate accuracies and the multi-temporal fusion method performs better than the multi-sensor fusion method. The results also showed that retrieval of sensor products is of higher accuracy than that of the spectral bands. The spatiotemporal changes of UHIs were indicative of an increasing trend over time. Of the surface biophysical parameters, the normalized difference built-up index exhibited the highest correlation with changes of normalized LST.