Evaluation of Grain Size as a Paleoclimate Proxy for Lake Level Change or Precipitation in a Large Lacustrine System: Flathead Lake, Montana

The use of grain size as a paleoclimatic proxy has been interpreted differently by various research groups. In some localities grain size fluctuations has been thought to represent precipitation changes connected with runoff intensity. Yet other studies have proposed that lake level variation or distance to sediment source is controlling grain size. In this study we sought to test each proxy interpretation to determine the strength of the data correlations and to calculate associated uncertainties with this methodology.

During the summers of 2000 and 2003, we recovered several piston cores from Flathead Lake, a large (496km2) open lake system in NW-Montana. One piston core from a central lake location was picked for high resolution grain size data analysis. We collected samples at every centimeter over the entire length (7.2m) of this core and analyzed grain size using a Malvern Mastersizer 2000 laser diffractometer.

Our initial results cannot support interpretations of grain size as a direct proxy for precipitation as proposed by others in similar large oligotrophic lacustrine systems with a single major sediment source. Correlations of grain size to lake level changes were stronger, however still contain uncertainty. In Flathead Lake this interpretation is complicated by the lake's complex post-glacial history of lake level changes due to down cutting of the terminal moraine, climatic changes, and rapid delta progradation. These components still need to be decomposed to improve the proxy value of grains size data. An unexpected major result has been how different grain size data collection methods can greatly impact data interpretation. We attempted to compare laser diffraction derived data with 20 year old settling tube data and found vastly different correlations and interpretations. Thus the study does show that interpretations based on different data collection methods are not directly comparable and need to be used with caution.