EWC'98 Homepage HOME | SESSIONS | SESSION PLAN | SEARCH | CONTACT

Possibility of medium-term forecast of strong winds

Marina Zdereva
West Siberian Regional Hydrometeorological Research Institute, Sovietskaya st. 30, Novosibirsk, Russia.
Tel. : +7-3832-222-530, Fax: +7-3832-226-347

Abstract:
The first model in Russia related to the forecast of dangerous or especially dangerous medium-term events has been developed. The methods of strong wind forecast available before had only a lead time of up to 24–36 hours in advance.

The territory involved in the model development includes areas of the Ural and of Siberia.

The main part of the investigations consisted in the study of the circulation conditions of dangerous events (DE) with respect to local physical-geographical peculiarities and in the formalisation of these conditions, in the search for informative predictors over the records and for possibilities of situations with or without DE separation. In the course of the work new quantitative characteristics of the atmosphere dynamics were developed which were previously not considered in the forecast methods.

The initial information for the DE forecasts are hydrodynamical forecasts of baric fields near the Earth and at the AT-500, of temperature at the AT-850, obtained from the European Centre for Medium-range Weather Forecast (ECMWF) in the GRID code, and the forecasts of extreme temperature according to the Institute’s model for five days.

The final version of models consists in the concordance of several branches including regression and discriminant analyses, the determination of conditions of the current synoptic situation, and the estimate of geostrophic wind values.

The forecast is produced in the form of 5 classes depending on the wind speed and on the degree of its spreading over the territory.

The results of DE forecasts depend on the quality of initial hydrodynamical products and on the peculiarities of the territories’ local circulation. The of author’s assessments and operational tests of the methods turned out to be very different depending on frequency of events in the regions. The most prominent failures were analysed, the main cause being gross errors of baric fields hydrodynamical forecasts. The main drawback of the models is the insufficient temporal (24 hours) and spatial (5° x 5°) resolution of the initial prognostic information. This is especially unfavourable for the processes of increased activity. The investigation of this problem resulted in partial solution performed with the help of interim baric fields build-up using movement trajectories of the air particles. Now some results of the strong wind forecast are used in operative synoptic work.