Spatial Analysis of Climatic Conditions for Cocoa Cultivation in West Africa: Assessing Current and Future Suitability

Aimed at deepening the understanding of the effects of climate variability on cocoa pro-duction in West Africa, conditions between 1959 and 2015 are analyzed based on new data sets of meteorological records, soil water content and evapotranspiration. Comparing the relationship of actual and potential evapotranspiration across the region, differences be-tween cocoa producing countries are found and discussed, affirming that cocoa producing regions in Ghana and Nigeria are the most restricted by water availability. Further, us-ing a machine learning approach (Maxent) the optimal climatic conditions for successful cultivation are determined and used in a model to assess climatic suitability. It is shown, that the precipitation during the driest month of the year is the most specific predictor of production suitability. The model overall predicts the area suitable for cocoa production with high accuracy (AUC = 0.983). Applying the model to climatic data from past years the average suitability is calculated for every county and year. Comparing its yearly vari-ability with reported yield commonalities are observed, but farming practices are found to affect the recorded yield more significantly. Further, the suitability time series shows large areas in Nigeria, Liberia and along the northern borders of the cocoa producing area being vulnerable to prolonged poor weather conditions. These areas are expected to profit most from measures aimed at increasing climate resilience.