147-19 X-Ray Fluorescence Scanning of Frozen Lacustrine Sediments: Application to High-Resolution Paleohydrologic Reconstruction

Session: Climate, Ocean and Environmental Changes Through Earth History: From Marine and Terrestrial Proxies to Model Assessments (Posters)

Poster Booth No.: 186

Presenting Author:

Abe Underhill

Authors:

Underhill, Abe¹, Steinman, Byron², Daniels, William³, Rodysill, Jessica R.⁴, Barré, Cole⁵, Brown, Rob⁶, Brown, Erik T.⁷

(1) University of Minnesota Duluth, Large Lakes Observatory, Duluth, MN, USA, (2) University of Minnesota Duluth, Large Lakes Observatory, Duluth, MN, USA, (3) University of Minnesota Duluth, Duluth, MN, USA, (4) USGS, Reston, VA, USA, (5) University of Minnesota Duluth, Duluth, MN, USA, (6) Large Lakes Observatory, Duluth, MN, USA, (7) University of Minnesota Duluth, Large Lakes Observatory, Duluth, MN, USA,

Abstract:

Comparison of lake sediment geochemistry with observational data is critical to reconstructing past environments over long timescales. Freeze coring is an excellent method for collecting undisturbed samples of lake surface sediments; however, frozen sediments are difficult to analyze using X-Ray Fluorescence (XRF) core scanning, as the sediment is water saturated and must remain frozen and relatively free of frost (which may attenuate signal) for the scan duration. Here, we present a method for XRF scanning frozen sediments, enhancing their use as high-resolution geochemical archives.

Frozen sediment cores were split into overlapping sections and placed in a custom-designed cold box to maintain core integrity during scans. Thaw tests in standard lab temperature and humidity conditions revealed that core sections can remain in ambient conditions for at least eight hours without degradation – exceeding the required duration for high-resolution XRF scans. Core surface topography proved most influential for accurate and reproducible scans. Anomalous measurements frequently occurred coincident with voids, divots, or other surface texture which were reduced by a surface scraping and cleaning regime. Frost typically formed on the core surface within the first hour after removal from cold storage, and a waiting period followed by removal of frost buildup on the surface improved scan reproducibility without the use of film.

Interpretation of the XRF data was most successful when normalizing by dividing the total counts for each element by the sum of the total counts of well measured elements at each depth interval, which were selected based on a correlation threshold (Pearson r ≥ 0.7) across three repeated scans of a single core section. This improved interpretation by excluding some matrix effects and density (water content) changes. Improvements of Pearson r values across three repeated scans averaged +0.08 and +0.02 for the uppermost and bottommost sediments, respectively. Data consistency between repeated scans was further improved via standardization of the normalized values using Z-scores. Sediment geochemistry to observational record correlation unlocks the possibility for streamflow and climate records to be extended to the past millennium, informing decision making, climate adaptation, and improving numerical and forecasting models of flood occurrence and magnitude.

Geological Society of America Abstracts with Program. Vol. 57, No. 6, 2025

doi: 10.1130/abs/2025AM-9165

© Copyright 2025 The Geological Society of America (GSA), all rights reserved.