A Comparison of Methods for Identifying and Quantifying Ice Rafted Debris on the Antarctic Margin

Abstract

Quantification of ice rafted debris (IRD) abundances in deep‐sea records is often used as a key proxy for identifying links between ice sheet instability and the oceanic overturning circulation. There currently exist multiple methods to determine IRD content in deep‐sea sediment cores. The preference for a given method is often determined by the accessibility to core material and destructive nature of some methods. While many studies have discussed the caveats between linking IRD to ice sheet dynamics, the uncertainties relating to the methodological approaches are often not considered in the interpretation or comparisons between different IRD datasets, particularly in the Antarctic. To address this, we compare three independent methodologies of obtaining IRD abundances and also discuss how different approaches will affect determinations of mass accumulation rates (MARs). The three methodologies we examine include: counting clasts >2 mm in x‐radiographs; the sieved weight percentage of the medium‐to‐coarse sand fraction (250 μm–2 mm); and volumetric estimates of the >125 μm sand fraction using laser diffraction particle size analysis. The x‐radiograph and sieve methods produced comparable results, while the laser particle size analysis, although showing comparable long‐term signals at most locations, in general has lower correlation to the other two methods and therefore a higher potential to obtain a noisy signal. We discuss the caveats associated with all methods, and emphasize that a combination of multiple methods should ideally be employed when assessing if a core is suitable for IRD study, and if sand or gravel fractions are accurately reflecting contents of IRD.

Unraveling the Deep‐Sea Sedimentary Record of Ice Sheet History

“…The rigorous comparison of methods commonly used to identify and quantify IRD provided by McKay et al. (2022) is especially timely since IRD is recognized as the fundamental proxy of ice sheet change preserved in deep sea sedimentary records. Recovery of high latitude paleoceanographic records from both hemispheres, especially those with glacial‐interglacial cycles representing warmer periods, has relied extensively on the technological and logistical expertise of IODP (i.e., Sánchez Montes et al., 2022).…”

“…McKay et al. (2022) acknowledge that although larger sample sizes are preferred the method chosen for quantification of IRD within a particular study may be dictated by sampling pressure resulting from a need to analyze several proxies (e.g., Müller et al., 2018) and limited available sediment.…”

“…McKay et al. (2022) compare three methodologies commonly used to determine IRD abundance within four cores collected from the Antarctic margin. These include: counting clasts >2 mm in x‐radiographs; the sieved weight percent of the medium‐to‐coarse sand fraction (250 μm–2 mm); and volumetric estimates of the >125 μm sand fraction using laser particle size analysis (LPSA).…”

“…However, the connection between deep sea IRD records and the specific retreat or collapse of the West Antarctic Ice Sheet requires recovery of drill core from areas proximal to the ice sheet (Escutia et al, 2019;Patterson et al, 2022). McKay et al (2022) emphasize the importance of facies analysis in addition to quantification of the IRD component in evaluating IRD MARs, which can appear to be increased by current winnowing at the seafloor. Recently, Andrews et al (2021) directly compared two measures of the IRD fraction (wt.% >2 mm and 250-2,000 μm) in a sediment core from near the expanded Iceland and East Greenland ice sheets.…”

Questions and clarity: insights from applying computational methods to paleoclimate archives

“…Based on the method employed and decisions made about how to define a Heinrich Layer, there may be significant experimental error in the reported thickness of layers. Indeed, an intercomparison of methods for measuring IRD abundance in cores from the Antarctic margin found that, though methods tend to represent the same trends of IRD, different methods could report IRD counts varying by a factor of two at a given depth in the core (McKay et al 2022).…”