Mesoscale meteorology models, such as the Weather Research and Forecasting (WRF) model, typically use land-use categorization data to specify surface characteristics by grid cells. In commonly used land use schemes, such as the 24-class USGS land-use classication system or the MODIS landcover product, cities are placed in single urban categories that are characterized by a high roughness length. Monin-Obunkov similarity theory is commonly used to simulate boundary layer dynamics based on surface energy and momentumfluxes which resulted in an over-prediction of wind speed in weather forecasts for urban areas. It is thus crucial to include the momentum drag effects of the buildings, including walls and roofs, in the prediction models. The WRF model contains several momentum flux parameterizations, including the single-layer urban canopy model, building effect parameterization (BEP) and building energy model (BEM). The BEM and BEP schemes are limited to the Mellor-Yamada-Janjic or BouLac planetary boundary layer (PBL) models and are computationally costly. In this study, a realistic multilayer urban approach is introduced into the Asymmetric Convective Model version 2 (ACM2) PBL component of the WRF model. Sensitivity tests are applied to a range of 31 urban morphological parameters in idealized simulations, and the impact of building morphology on atmospheric conditions is examined. The modified ACM2 model is implemented in two real cases in the forecast mode, resulting in significant improvements in wind speed prediction.
展开▼