Predictive data analytics for sustainable herding
We built a predictive dashboard to anticipate changes in water and grazing land availability in Senegal — to promote peace and stronger policy.
The story
Africa’s Sahel region is the semi-arid belt south of the Sahara Desert stretching from Senegal to Chad, where pastoralism and agriculture contribute to almost 90% of the economy. Increasing climate variability and the resulting onset of negative shocks, in conjunction with population growth, economic vulnerability, and social and political marginalization, are
Think-it’s role
Predictive analytics and model training.
-
Leverage satellite imagery from SentinelHub API to detect monthly water level changes — and train machine learning models for water surface availability: SARIMA AND RNN-LSTM.
-
Open-source platform development. Use Google Earth Engine to implement an application for detecting and embedding key factors in a dynamic dashboard —including NDVI, NDWI, and changes in temperature and population for the Sahel region



Why it mattered
We were able to affect two important areas:
Preventative Policymaking: Giving policymakers and NGOs data-driven insights to fuel more informed decision- making — like where to direct herding routes
Conflict prevention: Predicting and minimizing record-level incidents of violence and property damage between nomadic and agricultural communities.
Testimonials