Quantifying sheet wash erosion rates in a mountainous semi-arid basin using environmental radionuclides and a stream power model

2015 
Erosion rates and processes define how mountainous landscapes evolve. This study determines the range of erosion rates in a semi-arid landscape over decadal time spans and defines the dominant processes controlling variability in erosion rates. The varying topography and climatic regimes of the Xiying Basin (Qilian Shan Mountains, China) enables us to examine the relative roles of sheet wash versus rainsplash and the influence of vegetation on soil erosion and deposition. Soil erosion rates since 1954 were determined using Cs-137 along 21 transects at four sites with varying gradient, rainfall, and vegetation cover. The mean Cs-137 derived soil erosion rate similar to 0.42mm/a was consistent with the catchment level erosion rate derived from total sediment yield for a 44year record. However, there is considerable variability in Cs-137 erosion rates both between transects and along transects, perhaps reflecting variation not only in the effectiveness of individual processes but also in their relative roles. We compare the Cs-137-derived erosion rates with 1-D models for sediment flux that incorporate sheet wash and rainsplash processes, testing them over a previously untested 60year timescale. The variability in Cs-137 erosion rates along transects is best replicated by sheet wash dominated simulations, suggesting that this is the dominant erosion process in this semi-arid landscape. The functional form of the sheetwash model can also explain our observations that Cs-137 erosion rates decrease with upslope length (i.e. distance down slope) while its variability increases. However, sparsely vegetated sites, located in slightly drier locations, have higher erosion rates, and are not as accurately modeled as densely vegetated sites, suggesting that patchiness of vegetation introduces fine scale variability in erosion rates on these slopes. Copyright (c) 2015 John Wiley & Sons, Ltd.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    80
    References
    7
    Citations
    NaN
    KQI
    []