Regionalizing global climate models

Global climate models simulate the Earth's climate impressively at scales of continents and greater. At these scales, large-scale dynamics and physics largely define the climate. At spatial scales relevant to policy makers, and to impacts and adaptation, many other processes may affect regional and local climate and perhaps trigger teleconnections that provide significant feedbacks on the global climate. These processes include fire, irrigation, land cover change (including crops and urban landscapes), and the emissions of biogenic volatile organic compounds by vegetation. Many of these interact within the atmosphere via dynamical, physical, and chemical mechanisms that lead to boundary-layer feedbacks. It is unlikely that any of these processes have a significant global-scale impact on the Earth's climate in the sense that the amount of warming due to a doubling of well mixed greenhouse gases would change if these processes were explicitly represented in climate models. These phenomena are usually local in space (e.g. urban) or in time (e.g. fire) and probably do not provide the on-going and sustained forcing to affect the global climate. However, for most impacts and adaptation research it is the regional and local climate that defines climate risk. At these scales, processes missing in climate models can have a substantially larger local-scale impact than the additional radiative forcing due to increasing greenhouse gases. Thus, while climate models are well designed for global and continental scales they exclude a suite of important processes that are locally and/or regionally important. We review these missing processes and highlight the research required to resolve the representation of these regional-scale processes in climate models. We also discuss the experimental methodology required to rigorously determine whether these processes are restricted to a local or regional-scale role or whether they do trigger robust teleconnections that would demonstrate global-scale significance. Copyright (c) 2011 Royal Meteorological Society