Home Page (MVA) > Integrated Assessment Models (IAMs) and Resources > IAMs Thematic Guide
Thematic Guide to Integrated Assessment Modeling
[HOME] [PREVIOUS] [NEXT] [BOTTOM]
Resolution
A design question that arises repeatedly in integrated assessment projects, at least in those based on formal models, is spatial and temporal resolution. In integrating information across disciplines, integrated assessments must combine, reconcile, and propagate information that is available (and required) at different time and space scales. Procedures used to match and adjust resolutions involve substantial technical difficulties and are of potentially large significance for the ultimate utility and credibility of assessment projects.For example, the economic-energy models that project emissions typically operate in multi-year time steps on large-nation or regional political boundaries. Atmospheric chemistry takes place fast and on a small scale. Climate models run at fine temporal resolution, typically a 20-minute time step, but at moderately coarse spatial resolution, with grid cells of a few degrees (though spatial resolution is increasing quickly). Ecological impact studies typically need data at fine spatial resolutions, roughly a half degree, but their time resolution can vary greatly depending on the analytic approach taken, from daily (for some crop growth models) to monthly or seasonally.
Mismatch of resolutions can create problems even with the purely analytical exercise of providing an integrated description of the climate system. Making an assessment policy relevant for its audience can pose still more resolution problems, either because the decisions for which the audience is responsible are of coarser or finer scale than those the assessment studies--or do not appear at all in the assessment--or because the valuation and impact measures the audience cares about are coarser or finer than the assessment's outputs. Assessments that treat the European Community (EC) as a single region neither help European officials with their decisions allocating taxes or targets to member states, nor help national officials understand likely impacts of climate change for their countries. While aggregating fine-resolution results to coarser resolution typically poses no technical problems, going the other way is contentious and difficult.
The need to change resolution can arise in an integrated assessment at several points. The first is re-aggregating emissions modeled along political boundaries into latitude bands or cells, to provide locational input to atmospheric chemistry models. This is normally accomplished by judgmental allocations, which are normally assumed fixed and hence abstract away from whatever shifts in regional population and economic activity may occur over time.
The second, as discussed above, is downscaling results of climate models to spatial resolution fine enough to drive ecosystem studies. Of the two methods for increasing resolution of Global Circulation Models (GCMs), nested-grid modeling is too computationally intensive to be considered for integrated assessment studies. All integrated studies use some form of statistical downscaling, fitting over a climate model some observed pattern of weather from the historical record that agrees with the model results when averaged over the grid cell or over the dimension that is to be added. The basic question this method poses is whether the assumed correlation between large-scale and small-scale climate remains valid under the conditions of changed global climate that the model projects.
An alternative to downscaling is using analog climates, as in the MINK (Missouri-Iowa-Nebraska-Kansas region) and McKenzie Basin studies. The MINK study used climate records of the 1930s, which were available from local weather stations at essentially the resolution required for farm-scale impact studies. The McKenzie study is using a mixture of approaches, defining a judgmental climate scenario that combines analogs from past records with "locational analogs," imposing on one location the climate records of another presently warmer one. As with downscaling from GCM results, the basic question of this technique is the extent to which anomalous past climates, or climates from other places, can accurately represent a region's climate under changes in the global system.
The next section is Representing Policies.
