Key points are not available for this paper at this time.
A quantitative analysis is made of the range of operations for which conventional categorical weather forecasts are useful. It is shown that, because individual forecasts must of necessity be designed for only a small range of operations, their utility may be severely limited when the operating risks are much different from those to which the forecasts apply. This deficiency may be minimized, either by making categorical forecasts so that the optimum operating decision is provided for each user, or by providing an estimate of the probability of occurrence of critical weather and thus permitting the user to make his own operating decision.
J. C. Thompson (Sun,) studied this question.