Examination of mineral dust variability and linkages to climate and land-cover/land-use change over Asian drylands
MetadataShow full item record
Large uncertainties remain in estimating the anthropogenic fraction of mineral dust and the climatic impact of dust aerosol, partly due to a poor understanding of the dust source dynamics under the influence of climate variability and human-induced land-cover/land-use change (LCLUC). So far, the dust dynamics and linkage to climate and LCLUC in Central Asia have received little attention from the aerosol research community. This thesis comprises a comprehensive study of the dust dynamics in Central Asia focusing on 1) the seasonality of erosion threshold and dust emission affected by soil moisture, vegetation phenology and surface roughness, 2) the dust interannual variability and connections with large-scale climate variation (ENSO) through effects on the atmospheric circulation, precipitation, vegetation dynamics and drought, and 3) the impact of dust aerosol on surface radiative balance and photosynthetically active radiation, and possible effect on dryland ecosystems. A coupled dust model and multi-year ground and satellite observations of dust frequency, dust loading, and atmospheric and land conditions are used in this study. We find the threshold friction velocity significantly varies in space and time in response to soil moisture seasonality, surface roughness heterogeneity and vegetation phenology. Spring is associated a higher threshold friction velocity than summer, due to wetter soils and more vegetation cover. As a result, although more frequent strong winds occur during spring, spring dust emission is less than summer by 46.8% (or 60.4 Mt). Ignoring the dependence of the threshold friction velocity on the surface characteristics leads to biased spatial distribution and seasonality of dust emission. There is a strong linkage between dust and ENSO in Central Asia: La Nina years produce drought condition and enhance the dust activity. A decline in the strong wind frequency during 1999−2012 results in a decreasing trend in the modeled dust emission, at a rate of -7.81±2.73 Mt yr-1, as well as a decreasing trend in the ground observed dust frequency index, at a rate of -0.14±0.04%. We estimate that 58.4% of dust emission is caused by human activity during the 1999−2012 period. Our estimates suggest human plays an important role in the region’s dust budget through agriculture and water resource usage.