Modelling the impact of a Sub-Saharan metropolis (Kampala) on the local climate during specific meteorological conditions of a dry season

In order to build resilient cities in face of climate change in Sub-Saharan Africa, much is to be done to understand the impact of rapid and uncontrolled urbanization on the local climate in the region. Recent efforts by Brousse et al. (2019, 2020) demonstrated that using generic urban parameter information  derived out of Local Climate Zones (LCZ ; Stewart and Oke, 2012) maps created through the World Urban Database and Access Portal Tool framework (Ching et al. 2018) may be used to model the impact of Sub-Saharan African cities on their local climate – using the case of Kampala, the capital city of Uganda. These studies showed that despite the characteristic data scarcity on urban typologies that is present in Sub-Saharan Africa, LCZ could be used as a solution for modelling and studying the urban climates in the region.

Yet these conclusions were only obtained through the use of the bulk-level urban canopy model TERRA_URB, embedded in the COSMO-CLM regional climate model. We therefore test the applicability of a more complex urban canopy models – the Building Effect Parameterization coupled to the Building Energy Model (BEP-BEM) – over the region. To do so, we focus on short periods with specific meteorological conditions during the dry season spanning from December 2017 to February 2018. These are obtained through a k-means clustering over hourly weather measurements given by the automatic weather station located at the Makerere University, in the city-center of Kampala. Wind direction and speed, 2-meter air temperature, incoming short-wave radiation, precipitation, daily temperature range, 2-meter air relative humidity and near-surface pressure are used to depict 5 weather typologies (ie. clusters) during the dry season. We chose to keep only periods with 5 consecutive days of one weather typology, which results in three 5-day periods of distinct typology. We then run the model for these periods and evaluate its outputs against the state-of-the-art simulation by Brousse et al. (2020) as well as in-situ and satellite observations for certain meteorological variables. After that, we show the effect of the recent urbanization on the local climate for each of those three periods and relate it to the variability in urban heat.

This study is the first to model a tropical African city at 1 km horizontal resolution using the BEP-BEM model embedded in WRF. The latter could have major implications as more complex urban canopy models coupled to building energy models could shed light on the impact of the built environment on the livability of indoor and outdoor environments in these cities. Furthermore, insights could indeed be gained on the contribution of air conditioning heat fluxes to outdoor temperatures and the energetic consumption needed to keep indoor environments at an optimal temperature. Additionally, by resolving the urban environment in three dimensions, BEP-BEM could help increase our understanding of how specific urban planning and architectural adaptation strategies (like green or cool roofs, roof top solar panel, new building materials, urban greening etc.) may increase the citizens’ thermal comfort and reduce negative health impacts under specific weather conditions.