Fine-scale application of WRF-CAM5 during a dust storm episode over East Asia: Sensitivity to grid resolutions and aerosol activation parameterizations
Kai Wang,Yang Zhang,Xin Zhang,Jiwen Fan,L. Ruby Leung,Bo Zheng,Qiang Zhang,Kebin He
DOI: https://doi.org/10.1016/j.atmosenv.2017.12.014
IF: 5
2018-01-01
Atmospheric Environment
Abstract:An advanced online-coupled meteorology and chemistry model WRF-CAM5 has been applied to East Asia using triple-nested domains at different grid resolutions (i.e., 36-, 12-, and 4-km) to simulate a severe dust storm period in spring 2010. Analyses are performed to evaluate the model performance and investigate model sensitivity to different horizontal grid sizes and aerosol activation parameterizations and to examine aerosol-cloud interactions and their impacts on the air quality. A comprehensive model evaluation of the baseline simulations using the default Abdul-Razzak and Ghan (AG) aerosol activation scheme shows that the model can well predict major meteorological variables such as 2-m temperature (T2), water vapor mixing ratio (Q2), 10-m wind speed (WS10) and wind direction (WD10), and shortwave and longwave radiation across different resolutions with domain-average normalized mean biases typically within ±15%. The baseline simulations also show moderate biases for precipitation and moderate-to-large underpredictions for other major variables associated with aerosol-cloud interactions such as cloud droplet number concentration (CDNC), cloud optical thickness (COT), and cloud liquid water path (LWP) due to uncertainties or limitations in the aerosol-cloud treatments. The model performance is sensitive to grid resolutions, especially for surface meteorological variables such as T2, Q2, WS10, and WD10, with the performance generally improving at finer grid resolutions for those variables. Comparison of the sensitivity simulations with an alternative (i.e., the Fountoukis and Nenes (FN) series scheme) and the default (i.e., AG scheme) aerosol activation scheme shows that the former predicts larger values for cloud variables such as CDNC and COT across all grid resolutions and improves the overall domain-average model performance for many cloud/radiation variables and precipitation. Sensitivity simulations using the FN series scheme also have large impacts on radiations, T2, precipitation, and air quality (e.g., decreasing O3) through complex aerosol-radiation-cloud-chemistry feedbacks. The inclusion of adsorptive activation of dust particles in the FN series scheme has similar impacts on the meteorology and air quality but to lesser extent as compared to differences between the FN series and AG schemes. Compared to the overall differences between the FN series and AG schemes, impacts of adsorptive activation of dust particles can contribute significantly to the increase of total CDNC (∼45%) during dust storm events and indicate their importance in modulating regional climate over East Asia.