Modeling the Impact of Urban Climate on Vector-Borne Malaria in Sub-Saharan Africa Using COSMO-CLM—The Example of Kampala, Uganda

Malaria is mostly considered to be a rural disease. However, some studies documented its presence in urban areas (Robert et al. 2003, Hay et al. 2005, Castro et al. 2009, Machault et al. 2010) and Sub-Saharan Africa has been experiencing unprecedented rates of urban growth in the last decades (Seto et al. 2012). It is therefore necessary to better understand the impact of rapidly growing African cities on the local urban climate and thereby on climate-dependent diseases. In this work, we quantify the impact of urban climate on the presence of the malaria mosquito in the city of Kampala using the framework proposed by Brousse et al. (2019), which has proven to adequately address local data scarcity and allow for modelling the urban climate in Sub-Saharan Africa.

To this end, Local Climate Zones’ information (Stewart and Oke 2012), obtained following the WUDAPT framework (Ching et al. 2018), are fed them into the urban canopy model TERRA-URB (Wouters et al. 2016, Brousse et al. 2019) embedded in the COSMO-CLM model. The COSMO-CLM model is then run at a convection permitting scale of 2.8 km horizontal resolution, forced by ERA5 reanalysis data, before dynamically downscaling at 1 km over Kampala. Model outputs are evaluated against cloud-free land surface temperature, and precipitation measurements from satellite observations for the period 2010-2015.

Model outputs are used to calculate the dynamic temperature suitability index (TSI) proposed by Gething et al. (2011) to capture the temporal and seasonal evolution of the TSI. Brousse et al. (2019) already demonstrated using the static TSI – which doesn’t account for the temporal variability – that the urbanization of Kampala could lead to a 30% increase of the TSI in the city. In fact, this measure calculates the vectorial capacity of an environment for the development of malaria-infected mosquitoes out of air temperature. Thereby, hotter urban areas could have higher vectorial capacities. Hence, we also compute the effect of relative humidity on the survival of mosquitoes following Yamana and Eltahir (2013) in order to see how the dryer city may also influence the TSI. Both results of the suitability modelling are compared temporally and spatially.