The value of climate information: A decision‐analytic approach

A decision‐analytic approach to the problem of assessing the value of climate information is presented and an application of this methodology to a decision‐making problem involving the use of climate information is described. First, decision analysis is discussed briefly with emphasis on its informational aspects. Then a decision‐making situation is considered in which a farmer must decide which of two crops to plant in the face of uncertainty as to whether growing season precipitation will be above or below average. It is assumed that information concerning growing season precipitation is embodied in a climate forecast provided by a meteorologist and that this information constitutes the basis for the farmer's decision. Several probabilistic models for the climate forecasts are considered, including symmetric and asymmetric discrete distributions and a normal model. These models can provide reasonable descriptions of currently available climate forecasts. The analysis reveals that relatively uncertain climate information may be of value in some decision‐making situations, although the value of climate forecasts with skill levels commensurate with existing seasonal forecasts does not approach the value of perfect forecasts. Nevertheless, in the situations considered, once a minimum level of skill is achieved, further improvements in the forecasts result in steadily increasing benefits. Finally, it is indicated that decision analysis is also applicable to more complex situations in which climate information may play an important role. Copyright © 1983 John Wiley & Sons, Ltd

Duke Scholars

Author Robert L. Winkler Fuqua School of Business