Abstract
Food security interventions and policies need reliable estimates of crop production and the scope to enhance production on existing cropland. Here we assess the performance of two widely used ‘top-down’ gridded frameworks (Global Agro-ecological Zones and Agricultural Model Intercomparison and Improvement Project) versus an alternative ‘bottom-up’ approach (Global Yield Gap Atlas). The Global Yield Gap Atlas estimates extra production potential locally for a number of sites representing major breadbaskets and then upscales the results to larger spatial scales. We find that estimates from top-down frameworks are alarmingly unlikely, with estimated potential production being lower than current farm production at some locations. The consequences of using these coarse estimates to predict food security are illustrated by an example for sub-Saharan Africa, where using different approaches would lead to different prognoses about future cereal self-sufficiency. Our study shows that foresight about food security and associated agriculture research priority setting based on yield potential and yield gaps derived from top-down approaches are subject to a high degree of uncertainty and would benefit from incorporating estimates from bottom-up approaches.
Highlights
Food security interventions and policies need reliable estimates of crop production and the scope to enhance production on existing cropland
Yield potential estimated by AgMIP is 60% lower compared with Global Yield Gap Atlas (GYGA) across the four case studies (Fig. 2), which is consistent with the findings for other crop-producing regions (Supplementary Tables 2 and 3)
Agreement between GYGA and Global Agro-ecological Zones (GAEZ) was better at national and subcontinental scales, there were still large discrepancies between the two approaches, ranging from −50% to +30%. These differences were even larger at smaller spatial scales and for specific regions and crops, with GAEZ estimates differing from GYGA by −95% to 480% at local levels (Supplementary Tables 2 and 3)
Summary
Food security interventions and policies need reliable estimates of crop production and the scope to enhance production on existing cropland. Global assessments of future food security and land-use change published in high-profile journals have followed a ‘top-down’ approach that relies on crudely calibrated crop models and a gridded spatial framework to organize coarse data on climate, soil, and cropping systems to estimate yield potential and associated yield gaps[7,8,9,10] (Fig. 1 and Supplementary Table 1). An alternative to the use of top-down spatial frameworks is to follow a ‘bottom-up’ approach that estimates yield potential and yield gap for a number of sites explicitly chosen to best represent the spatial distribution of crop production area and upscales the yield potential estimated at those sites to larger spatial scales[13] (Fig. 1 and Supplementary Table 1). Yield potential estimation (3) Simulation of yield potential for each site using validated models and upscaling yield potential to climate zone (shown in map) and country levels (2) Retrieval of weather, soil and cropping system data from coarse, global databases (3) Simulation of yield potential for each grid based on generic model parameters
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
