Climate Models for Agricultural Impacts: Scales and Scenarios.

Much has been said about the need for high resolution (50 km or finer) climate scenarios for climate impacts assessments, including impacts on agriculture.  There have now been available for over a decade high resolution climate scenarios for application to impacts.  But the desire for high resolution must be viewed in the context of other major uncertainties regarding future climate.    These include the uncertainty regarding the future socio-economic, technological, and political development of the entire world, which will largely determine the emissions of heat-trapping greenhouse gases, and uncertainty in the response of the climate system to these emissions,  which is usually explored through the application of  the suite of global climate models developed at leading atmospheric science centers throughout the world.   Moreover, there are deeper uncertainties,  such as poorly understood climate processes (e.g., related to glaciers and ice sheets) that cannot be modeled,  as well as unknown processes at work.  It is also unclear on a pragmatic level, what the true added value of high resolution information is if it cannot be clearly established that information on climate change from regional climate models  is  truly more ‘credible’.  There has yet to be a study for example, that clearly demonstrates that superior adaptation planning can be developed from using high resolution climate scenarios as opposed to coarser (100s of kms) resolution ones from global climate models.  This paper explores the major aspects of uncertainty in climate scenarios and explores state of the art approaches to determining ‘credibility of scenarios’ and how scenarios can be applied in decision-making frameworks that emphasize robustness in the face of multiple uncertainties.