Аннотация:The probabilistic structure of significant mesoscale fluctuations in wind speed over Moscow is investigated. It is shown based on the data of long-term regular measurements at the Ostankino TV tower, as well as at other points of the megalopolis and in the surrounding countryside, that normalizing wind gusts makes it possible to obtain a unified probabilistic assessment of significant deviations of wind speed in the atmospheric boundary layer from the predicted values. A statistical analysis of long-term measurements of wind speed at six heights in Ostankino, at automatic meteorological stations at different points of the Moscow metropolis, special high-frequency measurements in the southwestern part of the capital, and remote sodar measurements of wind speed in urban and suburban areas showed that the distribution of normalized extreme deviations of wind speed has a general probabilistic structure and a universal distribution of maxima over the observation interval. Asymptotic parameters of the Gumbel distribution are given for the specified reference interval (3 h) and a sampling rate of automatic measurements (1 min). Although variations in the sampling rate and reference interval slightly change these parameters, they do not change the type of distribution. Conserving the shape of the distribution function of extreme values at different heights measured using different instruments at different observation points indicates that it is universal and allows us to estimate low probabilities of rare maximum wind gusts if the forecast of the average wind speed and energy of subgrid turbulent fluctuations are available from numerical models.
