Value of Predicting Environmental Variation in Fishery Management

Decision-makers are often concerned with forecasts of environmental variables. In accordance with this the quality of forecasts is often discussed. The interest in forecasts also show up in much modeling activity. However, the value of forecasts in terms of improved quality of decisions is not extensively studied. In this report we investigate the value of forecasts by the use of stochastic, dynamic optimization. The case is Northeast Arctic cod (Gadus Morhua) in the Barents Sea. We find that the value of ideal forecasts, when used to their full potential, is around 3 to 5 percent improvement in expected net present value. More realistic forecasts lead to improvements less than one percent. For practical purposes, it is only the forecast for the coming year that matters, long-term forecasts are of no use because the managed fishery is very flexible. These findings suggest that highly simplified forecasts can be used in models. The conclusions are likely to be somewhat sensitive to choice of model, a theme for further research. The project is financed by the Research Council of Norway, under the program Marine Resource Management, MARE.