Land Surface Temperature Trend and Its Drivers in East Africa

Land surface temperature (LST) is affected by surface-atmosphere interaction. Yet, the degree to which surface and atmospheric factors impact the magnitude of LST trend is not well established. Here, we used surface energy balance, boosted regression tree model, and satellite observation and reanalysis data to unravel the effects of surface factors (albedo, sensible heat, latent heat, and ground heat) as well as incoming radiation (shortwave and longwave) on LST trends in East Africa (EA). Our result showed that 11% of EA was affected by significant (p < 0.05) daytime annual LST trends, which exhibited both cooling of -0.19 K year(-1) (mainly in South Sudan and Sudan) and warming of 0.22 K year(-1) (mainly in Somalia and Kenya). The nighttime LST trends affected a large part of EA (31%) and were dominated by significant warming trend (0.06 K year(-1)). Influenced by contrasting daytime and nighttime LST trends, the diurnal LST range reduced in 15% of EA. The modeling result showed that latent heat flux (32%), incoming longwave radiation (30%), and shortwave radiation (23%) were stronger in explaining daytime LST trend. The effects of surface factors were stronger in both cooling and warming trends, whereas atmospheric factors had stronger control only on surface cooling trends. These results indicate the differential control of surface and atmospheric factors on warming and cooling trends, highlighting the importance of considering both factors for accurate evaluation of the LST trends in the future.