Gammon, P., Neville, L.A., Patterson, R.T., Savard, M.M., Swindles, G.T. 2015. A log-normal spectral analysis of inorganic grain size distributions from a Canadian boreal lake core: towards refining depositional process proxy data from high latitude lakes. Sedimentology
A high-resolution inorganic grain size dataset (701 samples) has been measured from a freeze core extracted from “ALE”, a boreal lake in northeastern Alberta, Canada. The grain size spectra are remarkably consistent throughout the core, exhibiting a structure comprising 6 (D1-D6) persistent grain size distributions (in a range from <1 to ~250 mm) plus a seventh (D7) medium sand distribution present in only 5 samples. Automated grain size spectral parametric peak-fitting was unsuccessful due to inconsistent solutions to highly similar spectra, and the migration of unambiguous modes to unrepresentative values. Constraining the modes of two of the distributions (D2, D5) produced parametric model fits that were statistically excellent (r2 > 0.999), consistent across samples with similar spectra, and consistent with the spectral structure exhibited across the whole dataset. Statistical analysis of the “constrained” solutions indicates that these models successfully extracted independent grain size populations. The multimodal ALE grain size spectra generate traditional grain size measures (e.g. mean grain size) that are composites of multiple grain size populations, implying their stratigraphic variation includes random errors due to variable contributions from different grain size populations. ALE is situated in a boreal wetland environment where inorganic sediment delivery is overwhelmingly dominated by surface overland flow transport during spring melt. The small size of ALE makes internal grain redistribution mechanisms inconsequential. The ALE grain size spectra is thus interpreted to reflect: 1) a bedload fraction (D4-D6 populations) transported during short-duration high discharge events during spring melt (i.e. relating to the intensity of the melt); and 2) a finer suspended load fraction (D1-D3 populations) representing suspended load material whose magnitude is controlled by the background spring-melt discharge (i.e. the volume of the spring melt). Stratigraphically both bedload and suspended load populations demonstrate short- and long-period, high and low amplitude oscillations, suggesting spring melt dynamics are themselves subject to external forcing factors that are cyclic at varying periodicities. The link established between the grain size spectra and spring melt dynamics has significant potential for generating long-term proxy records that better capture both the external forcing factors over spring melt in boreal systems, and the potential risks associated with spring melt hydrologic dynamics.
pdf coming soon