Simulating Fire Danger in West Africa under Climate Change Scenarios Using REMO-iMOVE-Fire Model

West Africa hosts one of the world's largest and most active fire regimes, shaped by the tight coupling of climate, vegetation, and human activity. This thesis investigates the current and future dynamics of West African fire using the regional coupled modelling framework REMO–iMOVE–FIRE, which links atmosphere, vegetation, and fire processes, and explicitly separates human and lightning ignitions. The model was evaluated against MODIS active-fire observations (2001–2018) and successfully reproduces the spatial extent of the savanna fire belt and its strong seasonal cycle. The analysis shows that present-day fire activity is overwhelmingly human-driven — when human ignitions are removed, fire activity nearly disappears. Future projections (2019–2045) under the high-emissions scenario RCP 8.5 indicate a widespread decline in fire numbers across the Sahelian, Sudanian, and northern Guinean belts, including Mali, Burkina Faso, Ghana, Benin, Togo, and Nigeria. This decline is driven by decreasing fuel loads, fewer high-danger days, and reduced human ignitions per unit area. However, protected areas such as Arli, Pendjari, and Gashaka-Gumti National Parks show potential increases in fire activity. These findings challenge the assumption that a warmer climate automatically increases fire. In West Africa, future fire regimes depend as much on people and fuel continuity as on climate, resulting in declining fire activity across much of the savanna belt.