Abstract
To advance understanding of the provenance of the silt and clay fraction (particles <0.063 mm in diameter) in the Taklamakan Desert, where dust events occur frequently, we have investigated the spatial distribution and geochemical characteristics of the silt and clay content of surface sediment in the region. The mean PM10 (particles <0.01 mm in diameter) and PM50 (particles <0.05 mm in diameter) content of the surface sediment of the mobile desert were <0.97% and <3.98%, respectively, which was lower than for the Gobi Desert (PM10 = 2.13–6.46%, PM50 = 7.94–20.86%) and the degraded land surrounding the Taklamakan Desert (PM10 = 3.71–13.70%, PM50 = 12.33–32.29%). These results indicate that the Taklamakan Desert is surrounded by potential sources of silt and clay. The relationships between TiO2 and Fe2O3, MgO, V, and Zr confirm that the geochemical composition of the silt and clay fraction have striking spatial differences in the interior of the Taklamakan Desert. Three regions with different silt and clay fraction provenances were identified: the Southeastern (I), Northeastern (II), and Western (III) deserts. Homogeneous geochemical characteristics of the silt and clay fractions from the Southeastern (I) Desert, Gobi Desert and degraded land surrounding the Taklamakan Desert, under prevailing wind conditions, means that the same sources of silt and clay fraction source existed among these areas. The significant differences in the geochemical characteristics of the silt and clay fraction from upwind degraded land, the Western (III) Desert and the Northeastern (II) Desert, indicate that they are sourced from different locations.
Acknowledgments
This work was supported by the Natural Science Foundation of China (40971006), the Desert Surface Processes and Mechanisms Program of the Hundred Talents Project of the Chinese Academy of Sciences, and the National Project of Scientific and Technical Supporting Program (No. 2006BAD26B 02-02).
Notes
n is the sample number.
See Figure for locations.
The means were significantly different among the subsets in all (P < 0.01, ANOVA).
a Robust Tests of Equality of Means (Brown-Forsythe tests).
n is the sample number.