At the large scale, and for a range of emission scenarios produced for the IPCC's Third Assessment Report (see Section 3.8), climate model results suggest that North America could warm at a rate of 1-3°C over the next century for a low-emissions case (B1). Warming could be as rapid as 3.5-7.5°C for the higher emission A2 case. Even the B1 case suggests substantially more warming over the next century than we have observed over the past 100 years. In addition, we know that the past century saw changes in temperature, precipitation, and other variables. Climate model predictions of precipitation remain highly uncertain. Many models suggest higher rainfall over North America accompanying warming in simulations of the IS92a emission scenario (Schimel et al., 1996). While some models suggest widespread and substantial increases in rainfall over most of North America, other models suggest a weaker increase in rainfall.

Understanding concurrent changes in regional temperature and precipitation is crucial. Warming with increasing precipitation is likely to increase plant growth, which may increase carbon storage (VEMAP Members, 1995) and may increase pest and pathogen invasion and expansion. Warming with a lesser increase or a decrease in precipitation could cause direct vegetation mortality and increase the risk of wildfire. Variability also plays a role. General increases in temperature and precipitation might increase plant growth, but occasional severe droughts would then maximize the chances of wildfire. Preliminary results from modeling of the United States suggest that projected changes in climate will cause very substantial changes to the distribution and productivity of ecosystems and to disturbance regimes (fire and drought probabilities). Subtropical conditions will extend further north into the United States, with accompanying changes to vegetation, hydrology, and the potential for disease. Changes at the Arctic border suggest changes to the forest-tundra transition region, losses of permafrost, and an altered growing season.

Concern about the spatial uncertainty of model-based climate scenarios has led to various attempts at downscaling global scenarios to regional scales. This can influence the results of impact studies, as illustrated by a recent case study of crop yields in the U.S. Great Plains (Mearns et al., 1999). Furthermore, published regional impact studies have used older global climate scenarios similar to the temperatures resulting from the new emissions cases, but these may not be directly comparable.

IPCC Homepage