Abstract
This report describes a method developed for routinely predicting the average characteristics of the forthcoming season from synoptic data gathered, processed, and analyzed on the macroscale. The method involves a combination of synoptic, statistical, and physical procedures and is at present partially, perhaps 60 percent, objective. The other 40 percent consists of experience in practical long-range forecasting, and thus involves imagination and ability to draw reasonable inferences from suggestive objective indications.
The method is illustrated by examples, and a verification of preliminary results is presented. In view of the short period the experiment has been in progress, the enormous complexity of the problem, and the fact that the research effort was largely skimmed off material and time primarily devoted to 5-day and 30-day outlooks, the conviction has been obtained that a greatly improved solution along the lines proposed is within grasp.
