干旱区生态环境知识资源中心

In this study, we present simulations of the global dust cycle for present day conditions using a new dust-atmosphere model based on the Non-hydrostatic Icosahedral Atmospheric Model (NICAM) coupled with the Spectral Radiation Transport Model for Aerosol Species (SPRINTARS), We focus on evaluations of the dust simulation with respect to emissions, depositions, surface concentrations, aerosol optical depths (AODs), and the dust-aerosol direct radiative effects (DREs). The sensitivities of the dust simulation to the meteorological fields are also investigated through with and without meteorological nudging. NICAM without meteorology nudging tends to systemically overestimate the 10 m wind speeds by approximately 30%-40%, whereas the daily magnitudes and variations in the 10 m wind speeds are both significantly improved with meteorological nudging, especially over the Sahara Desert. The estimated annual global mean dust emission flux, dust AOD, and dustaerosol shortwave DRE at the top of the atmosphere with meteorological nudging are 1463 Tg yr(-1), 0.033, and -1.3 Wm(-)(2), respectively. Due to the approximately 30%-40% overestimations of the 10 m wind speeds over the two major desert regions, the estimated annual global mean dust emission flux, AOD, and DRE without meteorological nudging are significantly greater than those with meteorological nudging. The overestimations of 10 m wind speeds and the associated dust emissions are mainly caused by the positive biases of wind speeds especially from surface to approximately 2 km and slightly affected by the temperature fields. The monthly variations in the dust depositions over the Atlantic and the surface dust concentrations over the Pacific are all better simulated with meteorological nudging. Compared to both the AERONET (Aerosol Robotics Network)and MODIS (Moderate-Resolution Imaging Spectroradiometer)- retrieved AODs, the simulated daily AOD variations are significantly improved with meteorological nudging, especially over the dust-aerosol dominated regions. The global and annual mean dust lifetime and size distribution, which are two critical factors for estimating dust radiative effects, are quite similar between the dynamic and nudged NICAMs. We therefore can use the dynamic model to understand climate-dust interactions in a global and annual scale. Furthermore, we can improve the model performances for some applications in regional and seasonal scales by meteorological nudging which probably cannot be achieved by just tuning the dust emission.