大气应急响应模拟中城市冠层风廓线参数化方法检验

城市冠层内风场的准确模拟或预报是突发性大气污染应急响应措施制定和实施的重要前提和基础。为了合理反映城市冠层的影响, 并满足应急响应时效性的要求, 将MacDonald(2000)提出的城市冠层内风廓线参数化方法耦合于中尺度气象模式MM5, 并利用2010年7月18日至8月6日北京325 m气象塔垂直观测资料进行验证。试验结果表明:(1)城市冠层参数化方法能够较好的模拟各种稳定度条件下冠层内风速廓线垂直变化, 中性、稳定和不稳定层结时的标准平均偏差分别为78%、12%、4%, 标准平均误差分别为78%、52%、21%。(2)城市冠层参数化方法能够较好的模拟冠层内实际风速变化情况, 虽然随高度增加模拟偏差增大, 但8、15、32、47 m高度的模拟风速与观测值依然十分接近, 标准平均偏差分别为2%、-26%、25%、60%, 标准平均误差分别为54%、46%、52%、73%。(3)与传统的Monin-Obukhov相似性边界层参数化方法相比, 城市冠层参数化方法明显提高了冠层风速的模拟能力。中性、稳定、不稳定层结时, Monin-Obukhov相似性边界层参数化方法的标准平均误差高达420%、176%、184%, 城市冠层参数化方法的标准平均误差减小至78%、52%、21%;冠层内8、15、32、47 m高度, Monin-Obukhov相似性边界层参数化方法的标准平均误差分别为283%、184%、227%、167%, 城市冠层参数化方法的标准平均误差减小至54%、46%、52%、73%。

Testing of Urban Canopy Wind Profile Parameterization for Atmospheric Emergency Response Modeling System

Simulation and forecasting of wind fields in the urban canopy layer (UCL) is important for developing and implementing measures for atmospheric emergency response. To reflect the impact of the UCL on wind fields and to fulfill rapid emergency response requirements, an urban canopy wind profile parameterization method (UCPM) proposed by MacDonald (2000) is coupled with the mesoscale model MM5. The performance of the UCPM is tested with observations from a 325-m meteorological tower in Beijing from 18 Jul to 6 Aug 2010. The results demonstrate that the UCPM accurately reproduces wind profiles in the UCL under neutral, stable, and unstable conditions with normalized mean bias of 78%, 12%, and 4%, respectively, and normalized mean error of 78%, 52%, and 21%, respectively. The UCPM shows excellent performances at heights of 8, 15, 32, and 47 m within the UCL with normalized mean bias of 2%, -26%, 25%, and 60%, respectively, and normalized mean error of 54%, 46%, 52%, and 73%, respectively. Comparison with the Monin-Obukhov similarity parameterization (MO) reveals that the UCPM significantly improves simulations of wind speeds in the UCL. Under neutral, stable, and unstable conditions, the normalized mean error of MO is 420%, 176%, and 184%, respectively, and UCPM is reduced to 78%, 52%, and 21%, respectively; at heights of 8, 15, 32, and 47 m within the UCL, the normalized mean error of MO is 283%, 184%, 227%, and 167%, respectively, and UCPM is reduced to 54%, 46%, 52%, and 73%, respectively.