Factors affecting the magnetic susceptibility of a loess-soil sequence, Barton County, Kansas, USA

To interpret climatic signatures from magnetic susceptibility (MS) in the Barton County loess-soil sequence, central Kansas, we tested the influence of particle size and carbonate concentration on magnetic susceptibility. First, sand, silt and clay were separated and MS was then measured on each of the three size fractions. To assess the influence of chemical treatments and carbonate on MS signals, the chemical agents used to remove carbonate, organic matter and free iron oxides were applied to another set of samples and MS was measured before and after each of the treatments. The results show that the bulk MS differentiates among stratigraphic units, but the environmental interpretation is far from straightforward. Soils have stronger MS signals than the loess units in the upper portion, while the carbonate-rich soils have weaker MS signals in the lower portion due to the MS dilution effect of carbonate. Sand MS peaks in Bignell Loess, Peoria loess and LL1 may suggest that in loess units fewer sand-sized magnetic minerals were changed by weathering processes into clay-sized magnetic minerals than in soils. A low bulk MS at the bottom of the reddish pedocomplex to the bottom of LS 1 is attributable to a combination of a high sand content and a high carbonate concentration. The silt MS peak in the Barton sand might have been derived from the parent materials. Comparing with mineralogic and chemical indices, clay MS might be closer to being a proxy of weathering intensity. Although clay MS dominates the bulk MS signals, silt MS and sand MS affect the bulk MS signals in two ways: (1) changes in the percentages of silt and sand, and (2) mineralogical variations in silt MS and sand MS from one stratigraphic unit to another. This study demonstrates that 40% of the variation in the bulk MS is accounted for by the clay MS and 56% of the variation by the silt MS and clay MS together. After adding the sand MS, 66% of the variation is accounted for. If diamagnetic carbonate is taken into consideration, 76% of the variation in bulk MS is accounted for by the MS of the three size fractions and the MS dilute effect of carbonate.