J3.4
Seasonal Forecasting of UK Winter Storminess
Steve E. George, University College London, Holmbury St. Mary, Surrey, United Kingdom; and M. A. Saunders
European windstorms caused damages of $1.8 billion per year (1990-1998). Skillful long range predictability of such events would benefit the risk management decision making of climate sensitive industries and emergency planners. We report a statistical model developed for the seasonal forecasting of UK windstorms. This is the first work of its type to have access to an extensive (in both time and space) UK Met Office dataset of hourly station wind gust data. A pre-season forecast for the December-March 1999/2000 period has been issued using climate data through September 1999. This predicts that the number of days with gusts reaching "strong gale" or "whole gale" levels are likely to be 20% above average in southeast and eastern England, and 0-10% above average in western England and Wales. The presentation will describe how the model is constructed, and address its level of predictive skill. On the latter issue,the results of hindcasting (1990-1998) indicate a mean error of 28% from actual (SE and E England), and 33% from actual (W England and Wales). A validation of the 1999/2000 winter forecast will also be provided. Suggestions will be made for the major physical interactions that enable such long-range predictions to be be made. Future developments, including the extension of the model to the European mainland, will be discussed.
Joint Session 3, Climate forecasting (Joint between 15th Conference on Probability and Statistics in the Atmospheric Sciences and 12th Conference on Applied Climatology)
Thursday, 11 May 2000, 8:40 AM-11:59 AM
Previous paper Next paper