Search

The purpose of the present paper is to review the application of numerical weather prediction techniques in the daily operational routine of the Hydrometeorological Service of the USSR. The theoretical aspects of numerical weather prediction models used have been discussed by the authors in original papers and will not, therefore, be repeated here.

REVIEW PAPER PRESENTED OCTOBER 5, 1967 AT THE XIVTH IUGG GENERAL ASSEMBLY LUCERNE, SWITZERLAND

This paper describes the preparation of a preliminary set of maps of gradient‐level wind climate over Canada and some neighbouring regions of the United States. The data for this study were drawn from twice‐daily records of radiosonde ascents at a network of ten Canadian upper air stations. The period covered by the data was 1960–69, and this was supplemented by statistical information on upper level winds at fourteen points in the United States. The analysis and subsequent mapping were carried out for 300 and 500 m above ground level and were based on a number of statistical models: the bivariate Gaussian distribution of wind velocities characterized by the vector mean and standard vector deviation, monthly Weibull distributions of wind speed and a new model of extreme annual wind occurrences developed from a consideration of wind as a stochastic process.