Albedo Retrieval From Multispectral Landsat 8 Observation in Urban Environment: Algorithm Validation by in situ Measurements

The surface albedo in urban environment represents one of the most influencing parameters in the formation of the Urban Heat Island (UHI). Its assessment is therefore strategic to undertake actions for the UHI mitigation: the increase of urban surface reflectivity is an example of cooling strategies increasingly proposed and developed. Spaceborne observations could give a significant contribution supplying albedo maps, and different retrieval algorithms estimating broadband albedo from narrowband reflectivities provided by satellite sensors were proposed and validated in literature. In this study, the accuracy of two well-known retrieval algorithms applied to Landsat 8 OLI reflectivities was assessed in a city of Central Italy (Perugia), selecting test sites having different land cover types. Since the root mean square error of the estimated albedo was quite high (0.07-0.08), a custom retrieval algorithm was developed to evaluate the potential improvement in the narrowband reflectivity-broadband albedo conversion for the case study. This analysis was made possible thanks to an in situ measurement campaign, carried out during the summer months of 2016. The selected sites, having quite homogenous features inside the urban environment, also demonstrated that the albedo values are not significantly affected by the hour of the day, if measurements are taken near noon (variation less than 7%). Spaceborne observations showed once again a powerful tool for monitoring large areas with a quite good spatial resolution (30 m), even if an improvement of the literature albedo retrieval algorithms is required and possible in an urban context.