Effect of urbanization on East African climate as simulated by coupled urban-climate model

Zeleke TT, Giorgi F, Diro GT, Zaitchik BF, Giuliani G, Ayal, D., Kassahun T, Sintayehu WD, Demissie T. 2023. Effect of urbanization on East African climate as simulated by coupled urban-climate model. Climate Services:31-100398.

This study examines the effect of urbanization on climate variability over East Africa. Seasonal trend of rainfall and temperature was analyzed using Mann-Kendall trend test and statistically significant rainfall trend is observed during spring (February-May) and summer (June-September) over northeast and spring/“bega”(October-January) seasons in southeastern regions of Ethiopia, thereby suggesting a seasonal shift of rainfall distribution. The temperature trend showed significant warming in the simulated field, except in central East Sudan, where there has been a significant decline. A numbers of idealized sensitivity experiments have been conducted with the Regional Climate Model (RegCM4.6) to investigate the contribution of urbanization to the East African region climate variability and trend. Model assessment against observed climate variables showed good performance in the simulation of spatial and temporal variability of regional climate variables. The results of the sensitivity experiment by prescribing different urban environments (tall building district (TBD), high density (HD), medium density (MD) and original land use) for the surface scheme (CLM4.5) reveal statistically significant impacts of urbanized surfaces on surface temperatures and precipitation due to variations in energy budget, local circulation and disturbance of hydro meteorological variables. It is noted that TBD urban environment has a higher impact on the local climate than other urban environments. Patterns of seasonal rainfall variability simulated using artificially urbanized land cover suggests involvement of complex interactions and is less similar to the observed rainfall trend, while surface temperature variability is significantly affected by local land-cover change and is very similar to the observed surface temperature trend.