Development of a real time prediction system for local wind assessment based on regional atmospheric models and its verification by the field tests

基于区域大气模式和微型风场模拟器，建立了一个考虑地形影响的真实的风场实时预报系统，并对该系统进行了风力发电评估。本研究提出了一种动态统计降尺度方法和理想化实现途径，该方法首先将水平分辨率为1 ~ 2km的中尺度大气模式嵌入全球模式中，对区域风场进行动态预报。其次，对区域风进行统计分析，以确定区域风气候。最后将区域风气候降尺度为水平分辨率为10 ~ 50 m的微尺度风气候。该方法的优点是能准确地预报山区的微尺度风，并能考虑海陆风、山谷风等局地环流的影响，用该方法预报的微尺度风在日本北部的塔比角进行了验证，结果与实测值吻合较好。塔皮灯塔和风电场的年平均风速预测误差分别为3.5%和8%。

A real time prediction system for local wind and wind power production assessment has been developed based on regional atmospheric models and micro wind simulator which can take the detailed topograpical effect into account. In this study, a Dynamical Statistical Downscaling Procedure and Idealizing and Realizing Approach were proposed.This method first dynamically predicts the regional wind by mesoscale atmospheric model with the horizontal resolution of 1 to 2 km which was embedded in the global model. Next, regional wind is statistically analyzed to determine regional wind climate. Finally, regional wind climate is downscaled to micro wind climate with horizontal resolution of 10 to 50m. The advantage of this procedure is to predict an accurate micro wind in mountainous area and to take the effect of local circulation such as sea-land breeze and mountain-valley wind into account.Micro wind predicted by this approach was verified at Tappi Cape, north of Japan and shows good agreement with the measurement. The prediction errors of annual mean wind speed at the Tappi lighthouse and the Wind Park are 3.5% and 8%, respectively.