Towards Early Response To Desert Locust Swarming In Eastern Africa By Estimating Timing Of Hatching
Elsevier
Download
Abstract: This academic study introduces a fuzzy logic-based ecological model developed to forecast the timing and location of desert locust hatching in eastern Africa, focusing on Sudan and Turkana County, Kenya. It incorporates environmental indicators like rainfall, soil moisture, temperature, and vegetation greenness to predict juvenile locust development stages. The research addresses the challenges of monitoring locust hatching in data-scarce and politically unstable regions by building an 81-rule inference system trained on multi-year field data from Sudan. The model demonstrates 82% prediction accuracy and correctly identifies seasonal hatching probabilities across Turkana?s subregions. Detailed spatial maps visualize monthly vulnerability, highlighting how precipitation and vegetation triggers align with locust development phases. The report underscores the limitations of conventional swarming forecasts and the advantages of ecological logic models in proactive pest control. It suggests operational integration into early warning systems to protect agro-pastoral communities and mitigate devastating economic and ecological consequences across ASAL zones.