Using Quartz as an Ice Rafted Debris (IRD) record

In addition to IP25, Axford et al 2011 (as mentioned in the previous post) also tested the presence of Quartz in marine cores as a proxy for ice rafting in the North Atlantic. Both of these proxies allow a reconstruction of ice abundance in the Northern Atlantic regions. Unlike IP25, however, Quartz is not a biological proxy, rather it is sometimes present as a 'foreign' material assumed to have been transported by floating ice (icebergs). Click 'Read more...' for a short summary of the method and a couple of examples.

Basically, the bedrock of Iceland is lacking in Quartz, but it is abundant in the rock of Greenland, Svalbard and other Arctic islands. Therefore, if Quartz is found in a core taken from a region where Quartz is not part of the geology, it is assumed to have been transported there from a region where it is abundant (apologies if this is completely obvious). In this case the 'transporter' is floating ice. By quantifying the fraction of this allochthonous material (from outside the local system) in the sediments through a core (using x-ray diffraction) a reconstruction of ice rafting to that area is possible. High levels of such Ice Rafted Debris (IRD) are indicative of an abundance of ice in the North Atlantic.

In a single core example, Moros et al (2006) quantified the quartz fraction in the sediment from a 25m marine core spanning 12Kyr taken from the Northern Iceland Ridge. Recent results compared favourably to ice records from the region and exhibited the regional patterns of the Medieval Warm Period (MWP) and Little Ice Age (LIA). A steady increase of Quartz was found from around 5/6Kyr to present preceded by a spike between 8 and 9Kyr.

Comparison between Moros et al 2006 (quartz %) and the median results from the multi-site Andrews et al 2009 study (Median qrtz wt%). Taken directly from Andrews et al 2009.

Single core studies, however, are prone to exhibit only local factors and may not be indicative of a pervasive regional pattern. For this reason, Andrews et al (2009) used similar methods on 16 core sites around Iceland and compared them with results from the Moros et al paper (see figure). They found a high correlation between the two (r=0.84) and the overall trend is present in both of the records, although the records do vary in their abundance suggesting that there is regional variation between the records as a result of current patterns. Also, they found no corellation between the results from this region and the ice records from Western Greenland and the Norwegian Sea showing regional variability in ice levels throughout the Holocene.

By incorporating multiple sites into an analysis, the authors reduced the likelihood of significant 'noise' in the data caused by unmeasurable effects of sediment entrainment and transport, an unknown more likely further down the cores as certainty of past ocean circulation reduces. Also, as with the IP25 proxy, the process is not completely understood as the exact sediment sources remain unknown.

This is a minor issue, however, and doesn't detract from the potential of Quartz as a general ice rafting indicator in this region. These examples also show the importance of using multiple sites in order to establish whether a proxy is showing a pervasive regional pattern or simply shows 'random' local variation.