Catastrophe forecasting: seeing "gray" among the "black boxes"
Purpose ? The purpose of this paper is to consider the problem of using "black-box" methods to forecast catastrophe events, and illustrate the value of independent peer review. Design/methodology/approach ? The problem with black-box catastrophe forecasts is the absence
of both extensive validation data and impartial peer review. These issues may be addressed by comparing black-box forecasts with a set of na´ve alternative forecasts provided by an independent party. To illustrate this approach, the historical hurricane forecasts of Dr William
M. Gray, professor at Colorado State University, are considered and a simple ARIMA analysis is offered as a na´ve alternative. Findings ? The analysis shows that Dr Gray's complex forecasting methodology does in fact provide reasonable forecasts, and may indeed offer value
beyond a na´ve alternative model. Originality/value ? The editorial identifies a major problem in catastrophe forecasting, and suggests one way to address this problem.