Spatiotemporal modeling of forest loss in Zambia using remote sensing and high-frequency climate metrics

Abstract

Drought is a global dilemma compounded by climate change, posing serious risks to populations reliant on natural resources. In south-central Africa, Zambia is faced with a high demand for rainfed agriculture to meet the food, energy, and income needs of its residents. This study investigates how drought influences forest degradation across Zambia's 116 districts over the 20 years from 2000 to 2020. Combining remote-sensed land cover and land-use change (LCLUC) data with high-resolution (5 km) drought indices, we employ fixed-effects regression modeling to examine the relationship between growing season drought and forest loss. Our findings reveal that drought events significantly increase deforestation rates, with the strongest effects observed up to five years after a drought, particularly in districts with lower water availability. These results suggest that forest degradation is not only a short-term coping strategy but also a delayed consequence of climate stress. By quantifying this relationship, we provide novel insight into the delayed, multi-year effects of drought on forest loss across Zambia's districts. These findings can improve public understanding of the broader impacts of drought and inform policies that support conservation efforts and the development of alternative energy sources to mitigate long-term forest loss.