USING SATELLITES AND MACHINE LEARNING TO ENHANCE AND PROTECT FOOD SECURITY IN AFRICA

When combined with data collected on the ground at farms, satellite Earth Observations (EO) data show what types of food crops are growing where and how they are doing as the growing season progresses. By comparing data from satellite images with reported crop yields from previous years, analysts can estimate how much food will likely be produced in a season. Combining satellite and ground data with climate information and weather forecasts enables us to forecast food production and get early warnings of potential crop failure from drought, floods, disease, and even pest infestations like the current 1 in 70-year locust invasion. This early warning gives organizations who respond to food insecurity more time to prepare and even mitigate shortages and famine altogether. Open data, machine learning, cloud computing and big data analytics are revolutionizing EO capabilities filling critical gaps in data even in the smallholder context. This presentation will summarize what satellite EO with machine learning offers today to support decisions that impact on smallholder farmers in sub-saharan Africa who face even more challenges and risk to agricultural productivity with climate change.