Abstract: We compare ensembles of water supply and demand projections from 10 global hydrological models and six global gridded crop models. These are produced as part of the Inter-Sectoral Impacts Model Intercomparison Project, with coordination from the Agricultural Model Intercomparison and Improvement Project, and driven by outputs of general circulation models run under representative concentration pathway 8.5 as part of the Fifth Coupled Model Intercomparison Project. Models project that direct climate impacts to maize, soybean, wheat, and rice involve losses of 400-1,400 Pcal (8-24% of present-day total) when CO2 fertilization effects are accounted for or 1,400-2,600 Pcal (24-43%) otherwise. Freshwater limitations in some irrigated regions (western United States; China; and West, South, and Central Asia) could necessitate the reversion of 20-60 Mha of cropland from irrigated to rainfed management by end-of-century, and a further loss of 600-2,900 Pcal of food production. In other regions (northern/eastern United States, parts of South America, much of Europe, and South East Asia) surplus water supply could in principle support a net increase in irrigation, although substantial investments in irrigation infrastructure would be required.

Abstract: Working with ensembles of crop models is a recent but important development in crop modeling which promises to lead to better uncertainty estimates for model projections and predictions, better predictions using the ensemble mean or median, and closer collaboration within the modeling community. There are numerous open questions about the best way to create and analyze such ensembles. Much can be learned from the field of climate modeling, given its much longer experience with ensembles. We draw on that experience to identify questions and make propositions that should help make ensemble modeling with crop models more rigorous and informative. The propositions include defining criteria for acceptance of models in a crop MME, exploring criteria for evaluating the degree of relatedness of models in a MME, studying the effect of number of models in the ensemble, development of a statistical model of model sampling, creation of a repository for MME results, studies of possible differential weighting of models in an ensemble, creation of single model ensembles based on sampling from the uncertainty distribution of parameter values or inputs specifically oriented toward uncertainty estimation, the creation of super ensembles that sample more than one source of uncertainty, the analysis of super ensemble results to obtain information on total uncertainty and the separate contributions of different sources of uncertainty and finally further investigation of the use of the multi-model mean or median as a predictor.