TDangarembizi/zimbabwe-drought-risk-analytics

Prototype drought risk forecasting and agricultural impact analysis for Zimbabwe using climate, satellite, and crop yield data to support early warning and evidence-based planning. # Zimbabwe Drought Risk Analytics ## Overview This project develops a prototype data-driven approach to forecasting drought risk and estimating agricultural impact in Zimbabwe. The focus is on early warning and decision support using publicly available climate, satellite, and crop yield data. The work represents an initial proof of concept, designed to be scalable during postgraduate study and applicable to real-world planning contexts. --- ## Problem Context Zimbabwe experiences recurring droughts that undermine food security, rural livelihoods, and national economic stability. Responses are often reactive due to limited integration of predictive analytics across climate and agricultural datasets. This project explores how data science methods can support earlier, evidence-based decision-making. --- ## Objectives - Analyse historical climate and vegetation trends relevant to drought conditions - Develop a prototype model to forecast drought risk at provincial level - Estimate potential impacts of drought on maize yields - Present outputs in a format suitable for non-technical decision-makers --- ## Scope This prototype focuses on: - Selected provinces within Zimbabwe - Maize as a priority staple crop - Monthly and seasonal climate indicators National scaling and integration of additional crops and regions are planned in later phases. --- ## Data Sources All data used in this project is publicly available, including: - Historical rainfall and temperature records - Satellite-derived vegetation indices such as NDVI - Crop yield statistics from international and national datasets - Climate indicators related to El Niño and La Niña patterns Where necessary, data is cleaned, aggregated, and aligned to ensure consistency across sources. --- ## Methodology The project follows a structured data science workflow: 1. Data collection and preprocessing 2. Exploratory data analysis to identify trends and anomalies 3. Feature engineering to capture seasonal and lag …