Impact of land-use and land-cover changes on mineral dust emission in Central and East Asia

The vast arid and semi-arid regions of Central and East Asia are one of the world's largest sources of atmospheric mineral dust. Over the past century human-induced changes in land-use resulted in the expansion of dust sources likely causing an additional dust burden (called anthropogenic dust). Given that atmospheric dust is an important climate forcing driver, reliable prediction of dust emission of both natural and anthropogenic origin is needed to improve our understanding of climate change in these regions. Our study addresses this issue by constraining the uncertainties in dust fluxes predicted by different dust production schemes considering individual dust sources with different topography, vegetation cover, and surface soil properties.

We use the PSU/NCAR MM5 mesoscale model in conjunction with a dust module DuMo to simulate dust emission and estimate the fraction of anthropogenic dust in the Aral Sea region and recently formed desert lands in Northeast China. Dust emission from the Central and East Asia regions for dust events in 2001, 2003 and 2004 was simulated at a range of model-grid spatial resolutions using dust production schemes of different complexity. In addition we explore the uncertainties in the anthropogenic dust emission by performing numerical experiments with varying land use/water masks corresponding to the historical land-cover/land-use changes occurred in Central and East Asia.

Our results indicate that the recently formed anthropogenic sources can play an important role at the local and regional scales, although their contribution over the entire region is relatively small compared to the natural arid and semi-arid dust sources in Central and East Asia. We also demonstrate that using different parameterizations for land surface processes, boundary layer physics, and dust production schemes can result in up to about 50% differences in predicted dust mass fluxes over the regions of interest.