中文摘要：

目前，耦合在WRF模式中的城市冠层方案包括单层冠层方案（uCM），多层冠层方案（BEP）以及考虑室内外大气能量交换的多层冠层方案（BEP＋BEM）。不同方案计算过程及参数设置不尽相同，这些方案能否适应高密度建筑物复杂下垫面的城市气象环境的模拟？建筑物形态参数对于城市气象因子模拟的敏感性如何？为回答上述问题，选取建筑物高密度城市——重庆为研究对象，以2006年重庆高温伏旱天气为背景，采用高分辨率（333m×333m）的地理信息系统数据替换WRF模式中默认的美国地质调查局（USGS）静态数据，对WRF中的3种城市冠层方案进行了模拟评估，并进行建筑物形态参数变化对气象因子影响的敏感试验。结果表明：（1）BEP＋BEM、BEP、UCM方案模拟值与城区内10个站点2m高气温观测值的均方差和平均误差分别为1．3、1．4、2．1和-0．5、-0．8、-1．4℃，单层冠层方案相对较差，BEP＋BEM方案最好。模拟值普遍略低于观测值。位于密集建筑物周边的2m高气温模拟值与自动站观测吻合相对较好，而中国国家基本站的观测值与模拟值相比，观测值偏低，靠近水体的模拟结果相对较差；（2）建筑物高度及密度参数的变化对城市近地层气温产生明显的影响，当建筑物高度增高时，由于短波遮蔽作用的影响，白天气温降低最大达到0．4℃，而夜间由于辐射截陷作用增强，气温明显升高，最大可达0．7℃。建筑物间距缩小，导致白天近地面气温降低，夜晚则升高，且夜间变化幅度更大；（3）当调整建筑物高度及密度参数分别为：高度20％（15m）＋60％（20m）＋20％（25m），间距20m时，城区内观测站点2m高气温观测值与模拟值的均方根误差从1．3℃减小到0．6℃，提高了模式的模拟性能。

英文摘要：

The three urban canopy schemes were coupled in the Advanced Research Weather Research andForecasting （WRF） model, including the single-layer Urban Canopy Model（UCM）, the multi-layer Building Environment Parameterization（BEP） scheme and the Building Environment Model based on BEP （BEP ＋ BEM）. Both the mechanism of urban land surface exchange and urban parameters setting are different from each other in these schemes, so it is very important to discuss whether the schemes and urban parameters setting are able to simulate the urban heat environment accurately for the urban with high densi- ty buildings and complex underlying surface in china. Also, evaluating the different scheme performance and investigating the building morphologic parameters sensitivity for urban heat environment simulation were worthy questions. In this paper three cases by selecting different urban canopy scheme in the WRF model were addressed for Chongqing City under continuous hot weather background in summer, and two group sensitive cases were designed to investigate the buildings＇ morphological pa rameters sensibility. The main results are as follows： （1）With the high resolution GIS data supported, the simulation result of BEP ＋ BEM scheme is the best. The RMSEs between observation and simulation of BEP ＋ BEM, BEP and UCM scheme are 1.3℃, 1.4℃, 2. 1℃, respectively. The simulated air temperature at 2 m was more consistent with the observation by auto- matic weather stations than by national basic stations. （2）Building height and density parameters have obviously influence on near surface air temperature. The building height increasing gives rise to the aspect ratio enlargement, then the short wave shielding is more distinct, which lead the air temperature at 2 m to reducing by the maximum of o. 4℃. While for the more heat storage in urban canopy owing to radiation trapping at night air temperature at 2 m increased significantly, up to 0.7℃. With the increase of buildings＂ density, the air tempera