Modeling Local vs. Global Dimensions of Food Security in Sub-Saharan Africa

Abstract:

Food security remains a daunting challenge in Sub-Saharan Africa despite dramatic efforts to foster innovation in the agricultural sector. Food security is complicated by a diversity of factors whose relative influence varies across scales, such as the nature of transportation infrastructure, the variety of agricultural practices, and the relative importance of food production versus food access. Efforts to model food security often focus on local-level dynamics (agricultural decision-making) or regional/coarse scale dynamics (e.g. GCM output + generalized equilibrium models of food trade) – both scales are of paramount importance to food security. Yet models of food security rarely span this scale divide.

We present work linking agent-based models of agricultural decision-making to regional and global dynamics of environmental change, food movement and virtual water trade in sub-Saharan Africa. Specifically we investigate the heterogeneity of environmental factors and agricultural decisions within the context of droughts of different duration and spatial extent. Drivers of meteorological drought manifest in agricultural drought through the complexity inherent in agricultural management. But efforts to model food security are often challenged by a lack of local-level empirical data to characterize the relationship between meteorological drought and agricultural drought. Our agent-based model is built using detailed information on household farm assets and individual farmer decisions, combined with crop yield estimated developed using the DSSAT cropping system model run with bias-corrected meteorological data. We then address food access through a analysis of food trade data given the increasing relevance of food movement to mitigate local and regional drought. We discuss the analytical challenges and opportunities in linking these cross-scale dynamics in food security modeling.