Forecasting and monitoring maize production using satellite imagery in Rwanda

Abstract

Agriculture is an important economic sector, employing over half of the workforce in Africa. The main threats that the sector faces are pests, plant diseases and climate change. Assessing and managing the risks posed by climate change for the agricultural sector requires access to high quality weather and crop yield data. Climate change is likely to have an impact on the variability of weather patterns in specific areas, which consequently disrupts the farming calendar. Extreme weather events such as drought have the potential to substantially reduce production. There are many programs across the globe that seek to increase agricultural productivity through specific farming practices, land consolidation and other agricultural strategies. However, their long-term success will also be affected by climate change. This study uses a quantitative approach to develop parsimonious models for forecasting and monitoring the performance of the maize crop while taking account of local rainfall estimated by satellite imagery. Understanding the influence of weather on past crop yields is the first step to quantifying the likely economic impact of climate change. The analysis also assesses the potential for improving the current system of crop land allocation by district in Rwanda.