Powerful computer programs are accurately simulating the uncertainty of climate change. These models underlie the growing business of climate risk, which in turn supports mitigation planning and the insurance industry. But climate risk models are usually large monolithic systems, which although internally consistent, cannot export statistically coherent representations of the underlying uncertainties. Because of this, the outputs are often single average risk scores resulting in the Flaw of Averages, a family of well documented mathematical errors. We propose open standards for conveying the results of climate models, based on the discipline of probability management, which will embed the uncertainty, including statistical dependence in auditable, cross-platform data. This will provide two major benefits. First, it will allow monolithic climate models to be disaggregated into manageable parts. Second, it will allow the results of disparate models of both hazards and impacts to be combined in numerous ways. In short, it will improve the measurement of environmental risk at any scale. Host Jack Russo and Professor Sam Savage of Stanford University discuss the use of probability management to create a new more Coherent Climate Calculus to effectively assess economic risks relating to climate change and toward the creation of entirely new climate finance marketplaces.
Want to learn more about probability management from Prof. Sam Savage? Check out his site www.probabilitymanagement.org