Meteorological observations of gust were carried out in Shonai Plain using X-band Doppler radar, anemometers, surface weather stations from December 2007. Based on the detected gusts accompanied with snowclouds, the meteorological features of gusts were revealed. Approximately 50 % of the detected gusts exceeded the wind speed of 20 m/s. Almost of the gusts were accompanied with convective radar echoes. Pressure jump of 0.3 hPa and temperature drop of 0.5 ℃ in average were recorded at the time of the gusts with convective echoes. Most of the gusts occurred within 10 km from the center of the echoes. These results implied that the drag force of graupel particle was an important factor of the downdraft (snowburst) and surface gusts. The layer of convective unstable up to 700 hPa was necessary for the developing of the convective clouds that brought the surface gusts as the environmental conditions.
In this study, a prediction method for strong gust based on onsite measurement and CFD was used which can take the effect of complex terrain into account. Predicted maximum gust shows good agreement with measurement. In addition, a time series based method for the prediction of the strong gust frequency in a control section was proposed. Predicted strong gust frequency agrees well with the measurement, while the conventional methods underestimate or overestimate the frequency. Finally, the proposed method was applied to the regulation of train operation, considering train speed, wind direction, and wind fences. The frequency of the regulation of the train operation at a 6km test section decreases from 0.16% to 0.14% when the train speed is reduced and decreases to 0.11% when the wind direction is considered, and further decreases to 0.08% when the wind fence is installed.