[A-AS11] Atmospheric Chemistry

[AAS11-02] High sensitivity of Asian dust emission, transport, deposition, and climate impacts to threshold friction velocity

*Kei Kawai¹, Hitoshi Matsui¹, Reiji Kimura², Masato Shinoda¹ (1.Nagoya University, 2.Tottori University)

Keywords:Asian dust, Kosa, global aerosol model

Asian dust (Kosa) is emitted from the surface into the atmosphere by strong winds in arid regions of East Asia (e.g., the Gobi and Taklimakan Deserts) and transported toward the North Pacific by the westerlies. In the source and downwind regions, Asian dust plays an important role not only in air quality and human health but also in climate through its interactions with solar radiation, clouds, and land and ocean biogeochemistry. The emission of Asian dust is controlled by many land surface parameters such as snow cover, soil moisture, and vegetation. In climate models, the effects of these parameters on dust emission are represented by the threshold friction velocity u_*t, and dust is emitted when friction velocity (wind speed on the surface) u_* exceeds u_*t. However, the treatment of u_*t has large uncertainties. In this study, we investigated the influences of u_*t treatment on estimating the emission, transport, and climate impacts of Asian dust using a global aerosol model (CAM5-chem/ATRAS2) and a land model (CLM4). In CLM4, u_*t is calculated from the threshold friction velocity over a smooth surface u_*t0 and factors dependent on land surface conditions. We performed two simulations in which u_*t0 was the default value of 0.23 m/s (BASE) and an observation-based value of 0.40 m/s (UST04) for the year 2012. The annual Asian dust emission for the year 2012 in the UST04 simulation was tuned to be the same as that in the BASE simulation. We compared observed and simulated dust event frequencies (the number of days when a dust event was observed per the total number of days during a given period) in East Asia. Compared with the BASE simulation, the UST04 simulation better reproduces the dust event frequencies observed in East Asia in spring and summer. The Asian dust emissions in the UST04 simulation increase by 31% in spring and decrease by 46% in summer and fall. Although the annual Asian dust emissions in both simulations are the same (289 Tg/yr), the annual mean atmospheric loading (the annual deposition amount) of Asian dust increases by 43% (49%) over the North Pacific and by 130% (73%) over the Arctic because of the relative increase in Asian dust emissions in spring. These results suggest that u_*t treatment further affects land and ocean biogeochemistry in the downwind regions as a source of iron and snow/ice-albedo feedback in the Arctic as a source of absorbing aerosols. The annual mean radiative forcing due to aerosol-radiation interactions in the UST04 simulation increases over the Gobi Desert and decreases from eastern China to Japan, and that due to aerosol-cloud interactions decreases in East Asia and the North Pacific. Our results demonstrate that uncertainties in u_*t treatment must be reduced, and not only the annual amounts but also the seasonal variations of Asian dust emissions should be well reproduced to improve simulations of the transport and deposition of Asian dust and to more accurately assess the roles of Asian dust in the Earth system (e.g., radiative balance, ecosystems, carbon cycle, and the cryosphere).