Exploring the Potentialities of Landsat 8 and Sentinel-2 Satellite Data for Estimating the Land Surface Albedo in Urban Areas Using GEE Platform

Albedo quantifies the capacity of a certain surface to reflect incident solar radiation. Therefore, this parameter is relevant in environmental and climate studies as it drives both the land surface energy balance and the interaction between surfaces and atmosphere. It can be estimated and monitored at different scales using Remote Sensing technique. However, its assessment is pretty difficult as factors, such as resultant map accuracy, processing time, and complexity of the algorithm used to retrieve it, should be considered. The goal of this paper is to develop a proper JavaScript code in Google Earth Engine cloud environment in order to estimate surface albedo using two different satellite data, Landsat 8 and Sentinel-2, over two different study areas: Bari (Southern Italy), and Berlin (Northeastern Germany). To achieve this purpose, Landsat 8 and Sentinel-2 images, acquired in close date, were processed in GEE environment by implementing an appropriate JavaScript code. After obtaining albedo maps over both investigated sites, the two algorithms’ performances, the Silva for Landsat 8 data and the Bonafoni for Sentinel-2 images, were statistically analyzed and compared. Furthermore, to investigate the outcomes deeply, statistics metrics were computed for different land cover classes also. UrbanAtlas provided by Copernicus was used to classify the whole case studies. Both approaches showed satisfying results albeit Landsat 8 algorithm provided higher mean values than the other one.