Uncertainty in integrated assessment models: modeling with MiniCAM 1.0

Human-initiated climate change remains one of the major science-based publi c policy issues facing the nations of the world. One feature of the issue t hat inhibits effective decisions is the ubiquity of uncertainty: many if no t most of the parameters of mathematical relationships that forecast emissi ons, atmospheric processes, impacts, and effectiveness of countermeasures a re uncertain. We have added a stochastic simulation capability to the commo nly used integrated assessment model MiniCAM 1.0 to analyze the sources of uncertainty and their relative importance and to help devise strategies for depicting and coping with uncertainty. The analysis shows that: (1) the pr ojected range of uncertainty in global temperature increase during the next century is greater than previously supposed, but the central tendency is n ot dramatically higher; (2) many of the major sources of uncertainty previo usly discussed in the literature remain significant, but not all; (3) an ad aptive policy of "act, then learn, then act" appears to offer better prospe cts for balancing uncertain costs and benefits of controlling greenhouse ga s emissions than do rigid precautionary measures; and (4) current targets f or atmospheric stabilization appear excessively ambitious. (C) 2000 Elsevie r Science Ltd. All rights reserved.