中文摘要：

激光测风雷达（Light Detection and Ranging,LiDAR）在大气边界层观测试验中的应用越来越广泛。测风精度对风资源评估非常重要。复杂地形流场的非均匀性导致激光测风雷达测风的误差,可能对相关的工程应用造成系统风险。文章给出了基于流体模型修正LiDAR测风误差的算法,并利用德国Fraunhofer IWES测试站同期LiDAR和风杯观测数据验证了算法的有效性。研究发现：（1）80 m高度各风向LiDAR的实测误差为-5.0%-2.5%,平均为1.2%;（2）稳态雷诺平均计算流体力学软件Greenwich CFD可用来修正复杂地形下LiDAR的观测误差,估算的分风向LiDAR测风误差和实测的LiDAR测风误差趋势是一致的,且模拟表明,Cosine型的误差形状与场址地形特征直接相关;（3）线性流体模型WAs P不能有效刻画坡度超过15°地形的流场特征,基于WAs P的复杂地形LiDAR测风误差修正是不可靠的。

英文摘要：

LiDAR （Light Detection and Ranging） becomes more and more popular in the observation of atmospheric boundary layer.The measurement accuracy of wind speed is very important for wind resource assessment.However,due to the inhomogeneity of wind flow over complex terrains wind speeds measured by LiDAR have some errors.It may cause systematic risk for relevant engineering applications.Basing on wind flow models,we here proposed an algorithm to reduce the LiDAR error.Fraunhofer IWES data from cup anemometers and a LiDAR during the same period are used to test our algorithm.Results show that our algorithm works well to reduce the LiDAR error of wind speeds over complex terrains.The measured LiDAR errors from observations at 80 m height above ground shows a variation from-5% to 2.5% with respect to wind direction sectors.The averaged error is about 1.2%.We also estimated the LiDAR errors by using the GreenWichCFD software which is based on the steady-state Reynolds-averaged Navier-Stokes CFD.It was found that the trend of estimated LiDAR errors basing on wind models is very similar with the measured LiDAR errors.The cosine shape of errors is related to the terrain characteristic of site.We also found that WAsP software which is based a linear wind flow model cannot estimate reliable LiDAR errors because it cannot simulate the wind field well when terrain slope is more than 15 degree at the inflow direction.