Multi-decadal analysis of high-resolution albedo changes induced by urbanization over contrasted Chinese cities based on Landsat data

Abstract

Surface albedo is a key parameter in the surface energy balance and has been identified as a primary essential climate variable (ECV). Variations in surface albedo can be used as a diagnostic tool for local climate change. This is particularly true in urban areas, where the impacts of land cover conversion due to increasing anthropogenic demands can be examined using surface albedo changes. Most of the previous studies of albedo in cities have relied on coarse-resolution datasets with short time spans and have disregarded continuous monitoring. In addition, it is still unclear which urbanization processes are involved and what effects they have on surface albedo over long time periods. This study aimed to identify the contribution of increasing urbanization to the regional climate by analyzing spatial and temporal changes in surface albedo. Assigning albedo values to land cover types is useful for determining the level of transformation and their impacts in various Chinese cities that underwent specific evolutions between 1986 and 2018. The Direct Estimation (DE) approach was modified to estimate the daily mean surface albedo at 30 m based on Landsat observations. It resulted root-mean-square errors (RMSEs) of less than 0.044 and bias about 0.006 between observations and model estimations. Such accuracy obtained after correcting the orbital drift of the Landsat satellite, was deemed satisfactory for detecting potential changes in albedo. Major findings are: 1) A notable trend was found over the past 33 years of 11 major Chinese cities, i.e. population about 10 million and more, with a general albedo increase from satellite observations. The higher resolution Landsat dataset showed a trend 3 times larger than the Global Land Surface Satellites (GLASS) product, which outlines the need for analyzing high resolution imagery in priority for reliable estimate of albedo over heterogeneous urban landscapes. 2) An increase in albedo infers a negative radiative forcing at an average rate of −2.771 W/m2 per decade, thereby indicating a cooling effect for most Chinese cities. 3) Changes in surface albedo were also closely linked to landscape transformation, clearly observed using the 30 m resolution of the Landsat data. 4) Throughout the study period, surface albedo exhibited a temporal U-shaped curve in built-up areas under development, with albedo peaks in old and newly built districts and a decrease in albedo between the two eras.