A land accounting model for IMPACT (with early results)

Understanding the global distribution of agricultural production provides valuable context for policymaking concerning development, wellbeing, and climate change. The IMPACT model generates broad regional aggregations of agricultural production and how much land would be needed, but a much more fine-grained picture would be helpful. The basic building blocks of such a companion system are presented here. Describing site specific land use and cropping choices as they are currently realized is a difficult proposition. Projecting them into the future is an even more challenging task. A key problem is that none of the simple “right†ways adequately mimic observed behavior. We present a mix of theoretical and heuristic approaches that bring together climate data, cost of access, and crop modeling results, to develop pixel level allocations of crops and natural lands consistent with economic simulations from the IMPACT model. We find that climate change is much more of a threat to natural land types than agricultural expansion into such locations.