Search

The constant building drag coefficient was replaced by a mechanical drag coefficient formulation in a multilayer urban canopy system coupled with the Rapid-refresh Multiscale Analysis and Prediction System to improve the wind speed prediction performance in complex urban areas. Although this formulation has been assessed in idealized urban areas by computational fluid dynamic simulations, the applicability of this formulation has not been sufficiently evaluated, especially in complex urban areas with inhomogeneously distributed tall buildings. The main objective of this study is to evaluate the performance of the coupled system in the Beijing urban area. The results show that the average wind speed model performance is improved by 40% in summer and 36% in winter by introducing a drag coefficient into the regional forecast model. [ABSTRACT FROM AUTHOR]

Realistic simulation of Madden–Julian Oscillation (MJO) propagation in coupled global climate models remains a common problem. In this study, the ability of 20 coupled models from Coupled Model Intercomparison Project Phase 5 (CMIP5) in simulating MJO is examined using dynamics-oriented diagnostics. The diagnostics focus on dynamic and thermodynamic structures of MJOs on three dimensions, which help to identify the shortcomings of models and evaluate whether they could reproduce MJO for the right reason. According to the simulation performance of the eastward propagation of MJO, the "good" models and "poor" models are detected. The dynamics-oriented diagnostics are further applied to the good models, poor models, and all models to establish a linkage between MJO simulation skill and their dynamic/thermodynamic structures. Results show that the simulations of good models have the following common features: (1) a horizontal zonal structural asymmetry in low-level zonal wind, upper-level diabatic heating and divergence; (2) a preceding eastward propagation of boundary layer moisture convergence; and (3) a rearward-tilted vertical structure of diabatic heating, equivalent potential temperature and available potential energy generation. The poor models that fail to capture these three-dimensional structures do not reproduce the eastward propagation of MJO. More than half of these 20 CMIP5 models still have difficulties in simulating these dynamic/thermodynamic structures. [ABSTRACT FROM AUTHOR]