In our thesis we proposed the statistical modelling concept to assess:
1.Which climatic factors are the most associated with malaria prevalence in Burundi.
2.Whether the forecasted increase in (global) temperature will result in increasing malaria transmission in Burundi.
3.Which amongst the well-known forecasting methods can provide a better forecasting of malaria cases in Burundi, when taking into account the effects of climatic factors.
4.Whether there exist spatial patterns of malaria which are explained by factors other than climate.
To achieve the objectives in our thesis, monthly data on malaria epidemiology and meteorology over 12 years for each province (of the area of Burundi) were collected from Burundi. Using these data, four studies are carried out. The first study proposes the analysis method based on a mathematical regression model to assess which variables significantly influence the malaria incidence in Burundi. The proposed modelling concept is based on both generalized linear models (GLM) and generalized additive mixed models (GAMM). The results obtained from these models reveal that malaria incidence in a given month in Burundi is strongly positively associated with the minimum temperature of the previous month. In contrast, it is found that rainfall and maximum temperature in a given month have a possible negative effect on malaria incidence of the same month. The second study proposes a Bayesian generalised additive model (GAM) to assess the impact of the predicted increase in temperature on malaria transmission in Burundi. The results obtained from the proposed model reveal that although malaria transmission is positively associated with minimum temperature, increasing temperature in Burundi will not result in increasing malaria transmission. In the third study we propose a hierarchical analysis approach to assess which method allows a better forecasting of malaria cases in Burundi when taking into account association between climatic factors and the disease. It is found that based on in-sample mean average percentage error (MAPE), the multiplicative exponential smoothing state space model with multiplicative error and seasonality leads to better forecasts (i.e. more accurate results). The fourth study proposes an analysis method based on the semi-parametric regression modelling of the dependence of malaria cases on spatial determinants and climatic covariates in Burundi. The results obtained from the modelling reveal some regional patterns of malaria that are related to factors other than climatic variables without being able to explain them.