Development of local wind prediction syste148 for improvement of wind resistant performance of the civil engineering structure in the mountainous region

点击翻译按钮获取中文摘要

A new database for the typhoon parameters, such as translation velocity, approach angle, central pressure difference, radius of maximum winds, annual occurrence rate, was established and a new stochastic model of the typhoon was proposed. Combining the data of isobars in weather chart and the observation data obtained at meteorological stations, the percentage of success in pressure field analysis raised from 13% to 66%. The new stochastic model for the statistical features of typhoon improved the accuracy of pressure parameters and translation velocity and the overestimate of expected wind speed for long-term return period by the conventional model. The extreme wind speeds at the six typical sites in Japan were simulated by Monte Carlo method. As a result, the value of annual maximum based on the new model showed good agreement with those from observational data. Comparing the expected annual maximum obtained by conventional method, it was found that the annual maximum wind speeds by the new model were reduced by 11% and 19% for the 100-year and 1000-year return period, respectively.Next, a new model for predicting strong wind in mountainous region during typhoon was proposed. The increase of wind speed on the left-hand side of typhoon due to the fast movement was reproduced. The local wind speed in during typhoon can be predicted by using the new wind field model and local wind prediction system developed in this study.Finally, the accuracy of new system was verified during the typhoon No. 21, which attacked Kanto plain on October 1^<st> 2002 and was compared with the observation data obtained at Suigo area, where the power transmission towers collapsed. As a result, the characteristic of strong wind in Suigo area and the detailed distribution of strong wind near the site was made clear.