Interface 2003

Predicting Damaging Climate Events: Methods, Examples, and Public Reaction
David W. Pierce, (Scripps Inst. Oceanography, Climate Research Division),, and
Tim P. Barnett, (Scripps Inst. Oceanography, Climate Research Division),


Climate events can cause significant damage to parts of the United States or world. The likelihood of occurrence for some damaging events are predictable months to years in advance, in cases where the forcing and physics of the phenomena are sufficiently well understood. Two examples of this will be presented. First, the massive El Nino of 1997/98 was predicted a year in advance, and generated intense public interest. This El Nino resulted in a significant amount of damage over portions of the United States, but also less severe than usual winter conditions over other parts. Our experience was that conveying the probabilistic nature of such events to the general public was difficult, and the results were often misinterpreted. Despite this, there is evidence that mitigation efforts (presumably spurred by the widespread media coverage) reduced damage from this event by a significant amount. Second, the Accelerated Climate Prediction Initiative (ACPI) project has used computer models of global warming to determine the changes in available water resources over the western United States for the next fifty years. The results suggest a high likelihood of large changes in the yearly water cycle, whose effects will be disastrous if not mitigated. The probabilistic nature of such forecasts will be discussed, and the reaction to these forecasts by state government authorities will be outlined.

Take me back to the main conference page.