Observation,Parameterization And Climate Simulation Of Dust Emission Processes

To fully understand the mechanism of dust emissions, uncertainties of mechanism in WRF-Chem and climatic characteristics of dust emissions, a field observation of dust devils was firstly conducted to estimate the vertical fluxes of dust emissions and gave an initial parameterization scheme. Then we used WRF-Chem to evaluate the uncertainties of dust emission processes. Finally, we gave a 20 years’climatic analysis of dust emissions with the global climate model CAM5.1. This paper’s contents and results are as follows:1) Observational study of formation mechanism, vertical structure and dust emission of dust devils over the Taklimakan DesertA field observation of dust devils was conducted at Xiaotang over the Taklimakan Desert (TD), China from July 7 to 14,2014. The measurements of dust devil opacity with the digital optical method (DOM) and the observed atmospheric boundary layer conditions were applied to investigate dust devils’ the formation mechanism, vertical structure and dust emissions. The critical conditions in the atmospheric boundary layer for dust devil formation were revealed with the land-air surface temperature difference of higher than 15℃, the enhanced momentum flux and sensible heat flux up to 0.54 kg m-1 s-2 and 327 W m-2, the weak vertical wind shear with the low wind shear index α<0.10 and the unstable stratification in the lower atmosphere. Based on observed dust opacities, it was identified that a typical dust devil was vertically structured with central updrafts and peripheral downdrafts of dust particles with the asymmetrically horizontal distribution of dust in a rotating dust column, and the vertical flux of near surface dust emissions was estimated in a range from 5.4×10-5 to 9.6×10-5 kg m-2 s-1 for a typical dust devil event over TD.2) Evaluation and analysis of the simulation uncertainty of dust emissionsBased on WRF-Chem, we simulated the dust emissions in dust storm processes in Taklimakan Desert by changing the soil moisture, the reanalysis data and parameterization schemes of dust emissions. The uncertainty of dust emissions was obvious. The major factors were the soil moisture, the reanalysis data and parameterization schemes of dust emissions. The effect of soil moisture on dust emission was significantly increased at the edge of the desert, with the difference of up to 500μg m-2 s-1 or more. The dust emissions were relatively high with Shao04 in Hotan and Korla, which was 2-3 times than GOCART. For NECP and ECWMF, the former was 3-9 times than the latter in Hotan and Tazhong.3) A simulated climatology of dust aerosol emissions over 1991—2010 and the influencing factors of atmospheric circulation over major deserts in the worldBased on the 20-year (1991～2010) simulation of dust aerosol emissions with the global climate model CAM5.1, the globally spatial and temporal variations of dust aerosol emissions over major deserts were climatologically characterized, and the atmospheric circulation factors influencing the dust emissions were analyzed. The results indicated that the annual amount of global dust aerosol emission was estimated with 1152±28 Mt in the 20-yr average, the dust aerosol sources were mainly concentrated in five major desert areas in North Africa, the Arabian Peninsula and Central Asia, East Asia, Australia and North America. As the world’s largest source of dust emissions, the deserts in North Africa were accounted for 61.8% of global dust emissions. The emissions over all deserts presented the distinct seasonal and interannual variations. Dust aerosol presented a seasonal cycle with strong emissions in spring and summer and weak emissions in autumn and winter. Compared to the seasonal fluctuations of dust emissions, the interannual variability was clearly weaker. Based on the significance levels of correlations between interannual variations of dust emissions and atmospheric circulation indices, the major climate factors influencing global and regional dust emissions over the deserts were determined:SOI (Southern Oscillation Index), AO (Arctic Oscillation), AAO (Antarctic Oscillation), AMO (Atlantic Multidecadal Oscillation), NP (North Pacific pattern) and WP (Western Pacific Index). The patterns of correlations between dust emissions and climate factors had obviously spatial distribution over the desert regions, and even the opposite patterns of correlations exist over a desert region. The La Nina- and El Nino-years in the ENSO cycle driven by air-sea exchange were connected with the more and less dust emissions in North Africa as well as the less and more dust emissions in Arabian Peninsula and Central Asia.