The Value of Frost Forecasting: A Bayesian Appraisal

Abstract

Abstract The economic value of frost forecasts is estimated under various assumptions concerning prior information, accuracy of forecasts, and the shape of the orchard operator's utility functions. The frost protection decision process is simulated in the context of Bayesian decision making under uncertainty. The averaged seasonal values estimated per day per acre were $@@‐@@5.39 for frost forecasts provided by the U.S. Weather Service, $@@‐@@8.57 for perfect frost forecasts, $@@‐@@4.73 for profit maximizers, and $@@‐@@191.39 for completely ignorant decision makers. The methodology used has general application to determination of economic value of information under conditions of uncertainty.