An analysis of future malaria intensity and climatic suitability in Uganda.

Abstract

Malaria is a large health threat on the African continent. The transmission of the parasite is highly dependent on precipitation and temperature levels. Relatively low temperatures in the highlands of Uganda can be inhibiting the current spread of malaria. Future climate change may therefore cause a shift in the climatic suitability for malaria transmission in this region. This study uses downscaled climate projections of temperature and precipitation values in Uganda from 2015 to 2099 generated by the MPI-ESM1-2-LR climate model. This data is used as input for a simplified version of the Liverpool Malaria Model. The model mathematically simulates several processes of the malaria transmission dynamics on a daily basis for every grid cell of the input data. The output consists of the daily basic reproductive rate (R0 ). The rate serves as a measure for the climatic suitability for and intensity of malaria transmission. Error margins for the climate projections are estimated through an ensemble created on a spatial subset of Uganda and are combined with the input data. The malaria model uncertainty is assessed through a sensitivity analysis using a one-at-a-time (OAT) method. Results show an increase in areas where, on average, malaria will spread (R0 > 1). This is most notable under climate change scenarios SSP3-7.0 (up to 24.1% of the area) and SSP 5-8.5 (up to 34.7% of the area). Several high-altitude locations in the west and east of Uganda show higher R0 values (R0 ≥ 2.5) starting from 2050. This is mainly caused by high levels of projected precipitation in the aforementioned scenarios. These locations also display the most notable increase in malaria transmission season length from 80 up to 120 days per year. The moment in time the season occurs appears stable. The error analysis warrants caution regarding the interpretation of these results, showing high errors for precipitation at high-altitude locations. Furthermore, the sensitivity analysis indicates a model that is highly sensitive to its parameter settings. Additional research needs to be carried out to calibrate the model parameters and to validate the outcomes of the current contribution.