MPAS-A Variable-Resolution Simulations for Summer Monsoon Over China: Comparison Between Global and Regional Configuration

摘要

Driven by ERA-Interim reanalysis, two climate simulations are conducted from April to August for 20 years (1998-2017) during the East Asia summer monsoon with Model for Prediction Across Scale-Atmosphere (MPAS-A). The global variable-resolution (MPAS-VR) and regional (MPAS-RCM) configurations are adopted under the same model framework with 92-25 km mesh refinement centered over Mainland China (MPAS-A experiments hereafter). MPAS-VR allows flexible interactions across scales. Constrained by the "reanalysis-based " lateral boundary conditions (LBCs), MPAS-RCM could well capture the evolution of large-scale fields and becomes a proper reference for evaluating MPAS-VR. The evaluations are conducted for precipitation, near-surface air temperature, and circulation features against observed climate. Two MPAS-A experiments present comparable results for the mean climatology and interannual variability except that MPAS-VR presents wet biases along the tropical ocean and underestimation in the Yangtze-Huaihe River basin. Compared with MPAS-VR, MPAS-RCM presents better daily variations and rain belts' evolution, which demonstrates the crucial effects of the LBCs for the limited-area model. The simulated circulation and moisture flux can explain most of the deviations. However, two MPAS-A experiments show consistent results for the extreme precipitation frequency and the extreme indices over Mainland China. With the same dynamic core and model physics, those climate features are similarly resolved by the two simulations. Thus, the 92-25 km MPAS-VR simulation is demonstrated as a promising tool for regional climate hindcast.In this work, we conduct model experiments by two configurations during the East Asia summer monsoon using Model for Prediction Across Scale-Atmosphere with refinement mesh centered over Mainland China. The global variable-resolution configuration (MPAS-VR hereafter) allows flexible interaction across global features to regional climate details. Constrained by the reanalysis (which blends simulated and observed results) around the boundary, the regional limited-area configuration with the truncated mesh (MPAS-RCM hereafter) is used for identifying the model's biases mostly related to its dynamic core and parameterization, which is a proper reference for MPAS-VR. We evaluate the model's performances against the observed climate for the near-surface air temperature and precipitation. Using MPAS-RCM as the criterion, MPAS-VR well captures the mean climatology, interannual variability, and extreme events, which demonstrates itself as a promising tool for regional climate details. The diverged circulation in the two experiments can explain most of the simulated biases. Extreme precipitation frequency and extreme indices over Mainland China suggest similar deficiencies in the two simulations due to a common model framework.