Meticulous reconstruction of the former extent of a glacier high in the mountains of New Zealand will help in interpreting global-scale climatic adjustments that occurred at the end of the last glaciation.
Both the Northern and Southern Hemispheres experienced phases of cooler climate that interrupted the general warming of Earth during late-glacial time, as ice sheets on the northern continents decayed. One of these cool phases, the Younger Dryas stadial (12,900 to 11,700 years ago), was triggered by huge volumes of cold, fresh meltwater entering the North Atlantic Ocean1. The result was a slowing-down of ocean circulation and a reduction of regional temperatures by up to 15 °C. In the south, the Younger Dryas was preceded by the Antarctic cold reversal (14,500 to 12,900 years ago), recorded in Antarctic ice cores2.
Understanding how far each cooling event propagated to lower latitudes is important for understanding how oceans distribute heat (and cold) across the planet. This can be achieved by dating the physical manifestations of such climate changes using preserved marine microfossils, plant pollen, insect faunas — or, as Kaplan et al. demonstrate (page 194 of this issue3), the bouldery moraine ridges deposited by former glaciers.
One problem is that terrestrial evidence of oceanographic change is fragmentary, linked to regional-scale climate and difficult to date with sufficient precision. Kaplan and colleagues' study from the Southern Alps of New Zealand is noteworthy in providing a well-constrained case study of a contemporary glacier advance. They have combined rigorous field investigations and state-of-the-art geochronology to resolve a long-standing debate4,5.
Kaplan et al.3 exploit the fact that small alpine glaciers are beautifully simple physical systems that respond sensitively to climate variability. The delicate balance between winter snow accumulation and summer melting determines the size and distribution of glaciers in a mountain range. Having reconstructed the location and topography of the former glacier from well-preserved moraine ridges (Fig. 1), they work out the sensitivity of glacier volume to temperature change by making some simple assumptions. They date moraines precisely, allowing a test of two competing hypotheses concerning the timing of glacier advance — did the advancing glacier deposit moraines in the earlier Antarctic cold reversal or in the later Younger Dryas stadial? It sounds a simple problem, but the authors adopt a rigorous methodology to address some intractable issues of age determination and interpretation.
Establishing the precise ages of boulder moraines in alpine environments, where opportunities for radiocarbon dating are few, has been problematic until recently. Various weathering-based techniques have been used6,7, sometimes with inconsistent or imprecise results. Kaplan et al. use cosmogenic beryllium isotopes (10Be) in quartz crystals as a measure of the exposure age of the rocks, enabling them to find out when moraine-building processes operated at a key period of global climatic transition. When the glacier dumped fresh boulders at its melting margins around 13,000 years ago, quartz crystals became exposed to cosmic-ray bombardment from deep space, initiating reactions to create the cosmogenic 10Be isotope. The authors have measured the 10Be concentration in 37 boulders to show, in essence, for how long these boulders have 'seen the sky'.
Although the use of cosmogenic isotopes in geomorphology has been revolutionary, the analytical uncertainties attached to each derived exposure age have proved stubbornly difficult to reduce. Principal among several sources of uncertainty is the rate of production of 10Be inside quartz crystals8. Typically, production-rate 'errors' are around ± 1,000 years, high enough to be fatal for attempts to separate the two late-glacial climate periods. Kaplan et al. have reduced their age uncertainties to only ± 300 years for individual analyses by using a local cosmogenic-beryllium production rate from a debris flow deposit that was dated independently by the traditional radiocarbon method9. This refinement is crucial and revealing: adoption of a less specific production rate would have obscured the true age of their moraines.
Kaplan et al. unambiguously resolve the competing hypotheses about the timing of glacier advance. Their principal contribution is to demonstrate a 'forcing' of the main glacier advance during the Antarctic cold reversal, peaking at 13,100 ± 500 years ago, a finding that reinforces other evidence from land and ocean for cooling of the southern climate at this time10,11. Even though a second moraine does date from the close of the Younger Dryas, the key point is that the glacier was retreating, almost uninterruptedly, through the period when dramatic cooling afflicted the North Atlantic region and caused the surrounding ice sheets and glaciers to readvance. Their climatic reconstruction shows a temperature difference of less than 1 °C between the two periods. In other words, the Younger Dryas had little influence in the southwestern Pacific Ocean — the main action had occurred 1,500 years earlier. The wider implication is that when ocean thermohaline circulation (that driven by density gradients) in the North Atlantic was suppressed in the Younger Dryas, the cooling did not propagate to southern mid-latitudes.
One might question how far we should extrapolate from the study of a single glacier, and the chain of reasoning from the hemispheric scale to the local and back again is a long one. Good science should leave us with a better set of questions than those we started with. Thus, we could ask whether the glacier retreat was solely due to the effect of temperature (rather than precipitation) on its mass balance, and what the relationship was between the local climate at the glacier and the regional climate of the time. Kaplan et al.3 justifiably claim to have found a missing piece in the jigsaw puzzle of late-glacial climate change: but how many such pieces are needed before the whole picture emerges? They contribute robust terrestrial evidence of late-glacial change, putting us in a better position to research why the climate changed, rather than debating the nature of the evidence. Their paper is a valuable chronological test, but it is not in itself evidence for a mechanism. Nevertheless, such finely resolved empirical evidence is vital for grounding our big ideas about climate change.
Broecker, W. S. Science 312, 1146–1148 (2006).
Jouzel, J. et al. Geophys. Res. Lett. 28, 3199–3202 (2001).
Kaplan, M. R. et al. Nature 467, 194–197 (2010).
Ivy-Ochs, S. et al. Geogr. Ann. 81A, 313–323 (1999).
Singer, C., Shulmeister, C. & McLea, B. Science 281, 812–814 (1998).
Birkeland, P. W. Geol. Soc. Am. Bull. 93, 433–449 (1982).
Winkler, S. NZ J. Geol. Geophys. 48, 105–116 (2005).
Balco, G., Stone, J. O., Lifton, N. A. & Dunai, T. J. Quat. Geochronol. 3, 174–195 (2008).
Putnam, A. E. et al. Quat. Geochronol. 5, 392–409 (2010).
Anderson, B. & Mackintosh, A. Geology 34, 121–124 (2006).
Pahnke, K. & Zahn, R. Science 307, 1741–1746 (2005).
About this article
New Zealand Geographer (2018)
Correlation of Late Quaternary moraines: impact of climate variability, glacier response, and chronological resolution
Quaternary Science Reviews (2012)