Out of tune: the dangers of aligning proxy archives Maarten Blaauw School of Geography, Archaeology and Palaeoecology, Queen’s University Belfast, BT7 1NN Belfast, UK article info Article history: Received 5 February 2010 Received in revised form 7 October 2010 Accepted 8 November 2010 Available online 15 December 2010 abstract Tuning is a widespread technique to combine, date and interpret multiple fossil proxy archives through aligning supposedly synchronous events between the archives. The approach will be reviewed by dis- cussing a number of literature examples, ranging from peat and tephra layers to orbital tuning and d 18 O series from marine and ice deposits. Potential problems will be highlighted such as the dangers of circular reasoning and unrecognised chronological uncertainties, and some solutions suggested. Fossil proxy research could become enhanced if tuning were approached in a more quantitative, reliable and objective way, and especially if individual proxy archives were non-tuned and kept on independent time- scales. Ó 2010 Elsevier Ltd. All rights reserved. Doubt is not a pleasant condition, but certainty is absurd. - Voltaire 1. Introduction The principles of superposition and lateral continuity can be used to deduce that distinct sediment layers were deposited simultaneously and originally extended horizontally in all direc- tions, even if now separated by for example a valley. Through aligning fossil proxy archives at wide spatial scales, supra-regional, hemispheric or even global pictures of past environmental changes can be reconstructed. The reasoning applied is that sediment boundaries or proxy events must have been produced by major climate or environmental events of such intensity and scale that they were registered in multiple regions and types of deposits in a (nearly) simultaneous manner. These same events can then be used as isochrons to date individual archives by aligning their proxy events to those in other, distant archives. Here I review this process of “tuning” or “wiggle-matching”. After presenting case studies of tuning from a range of published fossil proxy archives, potential problems are examined and alternative approaches suggested. All palaeoecological and palaeoclimatological proxy data come with a degree of uncertainty, in both the measurements of the proxies themselves and in their age estimates (Fig. 1). Therefore, whereas recent time series of perfectly measurable data can be represented by single “true” curves (Fig. 1a), historical environ- mental time series possess uncertainties in the proxy measurements (Fig. 1b), and palaeo data series have errors in both the proxies and the age-model (Fig. 1c). Even if these uncertainties are not generally depicted in fossil proxy diagrams (single curves as in Fig. 1a; Maher, 1972 , ; Blaauw et al., 2007), it is well-known that the data are uncertain and can thus be “moved around” in order to fit to certain hypotheses. For example, a major climate or environmental event might have caused environmental changes across a region, regis- tered as fossil proxy events in a range of sites (e.g., Gale, 2009), or the depth of a decline in Ulmus pollen in a single core can be assigned the age of the Elm decline as inferred from regional 14 C dated pollen archives. Measurement errors and internal variability will cause imperfect proxy expressions (e.g., slightly asynchronous) of these synchronous events. Therefore individual time series have some flexibility to be adapted, e.g., by aligning supposedly synchronous proxy events between archives, even if the raw data indicate asyn- chroneity (Figs. 2 and 3). The reverse can also happen, where time- transgressive events in diverse regions appear to have happened simultaneously owing to dating errors (Baillie, 1991; Fig. 3). As will become clear from the literature review below, there are several approaches to tuning. Firstly, one could apply no absolute dating information at all and assume synchroneity of proxy events with other archives (e.g., aligning pollen zones, tephra markers or d 18 O series). Secondly, one could date a proxy archive using a mix of absolute and relative tuning-based dating points (e.g., Cacho et al., 1999). Yet another approach is to date archives absolutely and independently to obtain an initial time-scale, after which the time- scale is adapted (preferably within its errors) to make proxy events align to those in other archives (e.g., Bond et al., 2001; Hoek and Bohncke, 2001; Neff et al., 2001; Burns et al., 2003; Charman et al., 2006). As discussed above, any difference in timing of dated events between proxy archives is then attributed to errors in E-mail address: maarten.blaauw@qub.ac.uk. Contents lists available at ScienceDirect Quaternary Science Reviews journal homepage: www.elsevier.com/locate/quascirev 0277-3791/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.quascirev.2010.11.012 Quaternary Science Reviews 36 (2012) 38e49