Sir

We agree with the overall thrust of your News story “Modellers deplore ‘short-termism’ on climate” (Nature 428, 593; 2004), with one exception. The US National Assessment of the Potential Consequences of Climate Variability and Change (USNA) — in which we were involved — did not attempt to provide regional or even national predictions of climate change, as was implied in the article. Nor, in our opinion, did it oversell the capacity of regional climate models to provide useful information.

The USNA report was explicit about its use of a range of approaches (mostly using projections from global climate models) to generate plausible scenarios of future climatic conditions that could then be used to explore the potential consequences of these scenarios for the environment, natural resources and people.

When the USNA performed regional analyses of global climate model projections, for specific areas of the United States, it did so in a way that cancelled out much of the systematic bias of the global models. This approach allowed the USNA to explore, for example, the responses of vegetation cover, agricultural production and water resources to the general character of the climate change expected during the twenty-first century.

Part of what the USNA learned and presented was a very clear description of the confidence the authors felt could be assigned to their conclusions. In most cases, the lack of true predictability of the climate results meant that the USNA presented many caveats to its findings.

Even a cursory reading of the USNA findings (http://www.usgcrp.gov/usgcrp/nacc) makes clear that they are not dependent on the regional details of the projected changes in climate.

For example, global warming is very likely to lead to reduced spring snowpack in the western mountains of the United States, reducing water resources; greater evaporation during the summer in the central states, lowering river and lake levels; a higher heat index that would endanger health in the humid eastern and southeastern regions; significant shifts in the landscape; rising sea level that will affect several low-lying regions, and so on.

We strongly agree that much more reliable regional climate simulations and analyses are needed. However, at present, as the News story makes clear, such simulations are more aspiration than reality.