GSA Annual Meeting in Denver, Colorado, USA - 2016

Paper No. 135-8
Presentation Time: 3:30 PM

CHEMICAL WEATHERING MODEL OF STREAM WATER CHEMISTRY IN A MEROKARST REGION DOES NOT REQUIRE LIMESTONE DISSOLUTION, KONZA PRAIRIE LTER SITE, NORTHEASTERN KANSAS


MACPHERSON, G.L.1, SULLIVAN, Pamela L.2 and JOHNSON, W.C.2, (1)Dept. of Geology, Univ of Kansas, 1475 Jayhawk Blvd, 120 Lindley Hall, Lawrence, KS 66045, (2)Department of Geography and Atmospheric Science, University of Kansas, 1425 Jayhawk Blvd, Lindley 210, Lawrence, KS 66045, glmac@ku.edu

We present results of PHREEQCi inverse modeling of one year of stream water chemistry at a limestone-mudstone lithology headwater stream, Konza Tallgrass Prairie LTER Site, northeastern Kansas, USA. The site hydrology indicates karst-like behavior, with rapid flow through bedrock joints and strong evidence of stream-aquifer interaction. Although the stream- and aquifer-water chemistry is typical of that in limestone aquifers, mass balance constraints suggest that the bedrock limestone is not the primary source of solutes, or the Permian-aged bedrock would have dissolved in 1-2 million years. Chemostatic to addition behavior of stream water K and Si suggest that interflow contributes significantly to stream water chemistry. The strong dominance of Ca on ion exchange sites, where illite is the most abundant soil clay mineral in the loess-based soils, as well as the likelihood of dust input to the region, also suggests most stream solutes may be the result of reactions within the soil, possibly with both in situ soil (leaving the soil carbonate-poor) and autochthonous dust. We subtracted Cl-weighted precipitation chemistry from the monthly stream water chemical analyses and used mass-balance (inverse) modeling to find reasonable combinations of mineral reactions to explain the water chemistry. The number of models depends strongly on the number of phases chosen to be available for chemical reactions, but, at one extreme, calcite (limestone) was not needed to account for the water chemistry. In that case, the mass of dust (recalculated with kaolinite component subtracted) needed to account for Ca and some of the other solutes was 50-80% of the average dust flux in the region. Because this is a grassland, a phytolith phase (SiO2 and K2O) was also tested and found to be a potential source and sink for Si and K. Although these inverse modeling results are non-unique, collectively they provide insight into mineral reactions that require minimal limestone-bedrock dissolution to account for the groundwater-fed stream chemistry.