Forecasting and Hindcasting the Maximum Combined Sea In Storms

Publisher Website
Google Scholar
Abstract
1. Abstract Forecasts of wind, waves and swells were made for a particular offshore operations area in the North Atlantic with accuracies of up to 97%. The operational forecast system, using a dynamic model, is sensitive to accurate measurements of wind velocity and the combined sea at the verification site. Some of the waves forecast with the system were "twenty year waves", that occurred more than once in four years, so that a statistical forecast system could not possibly have been used to anticipate them. The system uses the moving storm as the center of coordinates with as many details of the wind field as can be applied directly through measurement or indirectly through physical principles of the atmosphere in mesoscale. A simple circular storm model made from actual data for each case studied has resulted in hindcast accuracies never before achieved. This is shown to be a result of detailed wind information. The importance of detailed wind information for both forecasting and hindcasting cannot be overemphasized. 2. Introduction The Institute for Storm Research is responsible for forecasting the state of the sea, on a daily basis for specific clients with specific needs. A working system is utilized which not only accounts for the known parameters but also satisfies "real time" prediction requirements without sacrificing accuracy. The resulting predicted values are refined by the use of diagrams in order to account for the local effects on fetch length, shoaling and refraction. On a long term basis (five years of use), the system which is independent of location, has proven itself to be the best for offshore operations in shallow water. The Institute considers wave forecasting to be one of its primary products and is constantly analyzing, refining and updating its procedures to improve the product. A most effective way of refining forecasting techniques is to perform a "post mortem" analysis of selected storms and compare the results to actual observations. This is done by wave hindcasting and forecast evaluation. It is not sufficient to merelyrecalculate our forecasts using observed data. A re-analysis of the basic system must be made to insure that, while results may be satisfied through statistical verification, physical laws governing the sea's response to particular atmospheric conditions are not violated. Presented here, are the results of one such wave forecast evaluation study. The region of interest is the Grand Banks Area near Nova Scotia and Newfoundland particularly that area surrounding Sable Island. Two Computing Methods A method of computing significant wave heights has been developed to account for the proximity of the forecast point to a moving circular storm taking into account location, extent, intensity and movement.

This publication has 0 references indexed in Scilit: