The tsetse fly complex (Glossina spp.) transmits the parasite responsible for African Animal Trypanosomiasis (AAT), or nagana, which is the most economically important livestock disease in Africa. The challenge of the last decades was to design programs that could sustainably control fly populations in different regions of the continent. Given the contrasted outcomes of these programs, there is still a need for a better understanding of the spatio-temporal dynamics of this vector. Mathematical models and computer-based simulations are relevant to assess which control measures should be used and when, accounting for the ecological complexity of the target pest and territorial specificities of the controlled area. They provide a useful tool, complementary to field observations and experiments, to suggest efficient vector management strategies. We developed a deterministic and mechanistic spatio-temporal model of the population dynamics of tsetse flies, structured by sex and age (pupae, teneral and non teneral adults). Temperature and fly density influenced the life-cycle, while spatial diffusion depended on density and relative quality of neighbouring locations. We applied the model on populations of Glossina palpalis gambiensis in the Niayes area of Senegal, for which biological and landscape data were available. The landscape was divided into 250m x 250m cells of heterogeneous carrying capacity, estimated by habitat suitability models. We transformed observed temperatures into “perceived” ones, to account for micro-environments where flies live. Dispersal, mortality, and development rates were calibrated on laboratory data, experts’ opinions and literature. The sensitivity analysis of the model identified the biological and environmental parameters influencing the most population dynamics. We showed that the mortality and development of adult females, along with temperature, were the key drivers of population persistence. Our predictions suggested that combining techniques to both increase mortality and decrease fecundity would be optimal to eradicate tsetse flies in targeted zones. Sequential aerosol technique (SAT), traps and targets (TT) and insecticide-treated livestock (ITL) increase daily mortality rates, whereas the sterile insect technique (SIT), by preventing egg-laying, slows down the development rate of the population. Furthermore, our results highlighted the need for more biological insights to achieve accurate model predictions. Additional field work and experiments are necessary to better infer the relationship between adult mortality and temperature, as well as differences between temperatures from weather stations and temperatures in tsetse fly resting places.