Striking aeolian landforms characterize loess landscapes of the Great Plains and Upper Mississippi Valley, USA, shaped in Late Pleistocene environments with many characteristics of modern deserts including large active dunefields. Similar aeolian morphodynamics are evident in northern China and the Columbia Basin, USA, and are clearly important for interpreting the paleoenvironmental record of loess. Four zones spanning the upwind margin of thick loess can be defined from landforms and surficial deposits. From upwind to downwind, they are: A) A largely loess-free landscape, with patchy to continuous aeolian sand mantling bedrock. B) Patchy loess deposits, often streamlined and potentially wind-aligned, intermingled with dunes and sand sheets; interbedding of loess and sand may be common. C) Thick, coarse loess with an abrupt upwind edge, with troughs, yardang-like ridges, and/or wind-aligned scarps recording large-scale wind erosion. D) Thinner, finer loess with little evidence of post-depositional wind erosion. The degree of development and spatial scale of these zones varies among the loess regions we studied. To explain this zonation we emphasize controls on re-entrainment of loess by the wind after initial deposition, across gradients of climate and vegetation. The role of saltating sand in dust entrainment through abrasion of fine materials is well known. Using the Portable In situ Wind Erosion Laboratory (PI-SWERL), we recently demonstrated that unvegetated Great Plains loess can also be directly entrained under wind conditions common in the region today (Sweeney et al., 2011, GSA Abstracts with Programs, Vol. 43, No. 5, p. 251). Rainfall-induced crusts largely prevent direct entrainment in fine loess, but appear less effective in coarse loess. We propose that in zone A, any loess deposited was both abraded by saltating sand and directly re-entrained, so none accumulated. Sparse vegetation in this zone was primarily an effect of climate, but the resulting aeolian sand transport could itself have reduced vegetation growth and weakened crusts on loess patches. Loess re-entrainment was less effective in zone B, because of more frequent vegetation cover or other limits on aeolian sand transport. The abrupt transition to thick loess accumulation in Zone C is often associated with a major barrier to sand transport, e.g. an incised stream valley. In the absence of saltators, much of the loess in zone C was stabilized by vegetation and/or crust formation. However, preferential deposition of coarse loess in zone C resulted in weaker crusts and lower plant-available water, so in this zone the loess was still susceptible to large-scale wind erosion. In contrast, finer loess was deposited downwind in zone D, so aeolian re-entrainment was minimal. The large troughs and scarps in zone C would certainly have affected near-surface wind flow and the patterns of wind erosion and deposition. We have not explored this feedback in detail, but there is topographic evidence for enhanced loess deposition rates related to flow separation atop steep escarpments such as the sides of troughs. Enhanced deposition on loess escarpments is responsible for especially high-resolution Holocene loess records in the Great Plains.
|Original language||English (US)|
|Title of host publication||American Geophysical Union Fall Meeting|
|Place of Publication||Washington, DC|
|Publisher||American Geophysical Union|
|State||Published - 2012|