North-South glacier asynchrony during the Holocene

11 10 2012

Well-constrained reconstruction of glacier activity during the Holocene epoch is difficult for a number of reasons, principally:

  • (palaeo)glaciers are often located in harsh, high-altitude environments, which often lack sufficient productivity for radiocarbon dating;
  • until recently, surface exposure dating techniques have lacked the precision to resolve the age of young landform formation (although see this study and this study);
  • in the northern hemisphere at least, glaciers were at their maximum Holocene position relatively recently during the ‘Little Ice Age’ approx. 200-300 years ago, thereby removing any geomorphological record of previous glacier positions upvalley.

Holocene moraines denote former glacier position in Greenland (Humlum, 2000)

For these reasons alternative, indirect proxies of past glacier activity have often been sought. In a recent issue of Quaternary Science Reviews, Canadian-based authors Maurer et al. present a record of Holocene glacier fluctuation in British Columbia, Canada, using a sedimentary record from small mountain lake.

The reason this study caught my attention was the rather advantageous geomorphological setting, which allows the story of past glaciation to be told. The lake is located in a catchment which is currently unglaciated, but during times of more extensive glaciation the tongue of the nearby valley glacier extended into the lake catchment, providing inflow via glacial meltwater. When a glacier is present in the lake catchment the lake receives much coarser grained sediment than normal. Recognising this, Maurer et al. describe the past fluctuation of this glacier through the interpretation of a series of 3 lake sediment cores. They informally term the lake ‘On-Off lake’, presumably to reflect the numerous switches between glacial and non-glacial inputs. Radiocarbon dating of organic layers in the sediment cores and of tree remains in the glacier forefield provided the temporal constraint for identified glacier fluctuations.

Example of changing sediment types in sedimentary cores (source:

It was found that during the first half of the Holocene, from 10000-5000yrs before present, the local glacier was not extensive enough to provide input to On-Off lake. Tree trunks showing evidence of having been sheared by an advancing glacier were dated to approximatey 5000yrs old, representing the first evidence for local glacier advance during the Holocene. Between approximately 2750 years ago and the present day, the glacier front advanced and retreated over the On-Off lake drainage divide several times, as indicated by the switches between glacial and non-glacial sediments in the lake cores. Maurer et al. use historical photographs to supplement their geological record of Holocene glacier behaviour to the present day.

This study corroborates other records of northern hemisphere glacier activity, which show reduced glacier extent similar to / less than that of the present day during the early Holocene, followed by glacier expansion in the latter half of the epoch, culminating in the Little Ice Age maxima 200-300years ago. Recent studies in the southern hemisphere appear to show the opposite behaviour, with glaciers at their maximum Holocene positions in the middle of the epoch (approximately 6000years ago) and have undergone overall retreat ever since. These studies (link and link) suggest that this asynchronous behaviour between hemispheres is connected and caused by the changing position of the intertropical convergence zone (ITCZ). This climatic boundary shifts northwards and southwards over various timescales, predominantly in accordance with changes in the Earth’s orbit of the sun. During the Holocene the ITCZ has slowly shifted southwards, increasing the proximity of southern hemisphere glaciers, such as those in New Zealand’s Southern Alps, to warm, tropical airflows. At the same time, this tropical influence is therefore reduced in the northern hemisphere mid-high latitudes. Glaciers are highly sensitive to changes in atmospheric temperature, therefore it is suggested that these changes in ITCZ position and the respective control on regional air temperature are responsible for the differences in Holocene glacier behaviour between the hemispheres.

ITCZ temperature influence cartoon (source:

These findings are interesting as they suggest that the controls on glacier extent can switch between regionally-dominant (such as that shown above for the Holocene) or globally-dominant (such as during the last glacial maximum approximately 210000years ago, when glaciers around the world advanced synchronously). As this study (link) points out, since the industrial revolution c.1850 AD, the majority of the worlds glaciers have been retreating in synchrony, likely due to rising temperatures in response to increased global greenhouse gas levels. Is this the case? If so, was global greenhouse gas concentration controlling the synchronous glacier behaviour during the last glacial maximum? And, what causes the switches between regional/global dominant controls on glacier behaviour? Further addition of well-dated, palaeoglacier and palaeoclimate records from around the globe will help to test hypotheses that aim to resolve these questions.

IHQS research paper summary: Golledge et al (2012). Last Glacial Maximum climate in New Zealand inferred from a modelled Southern Alps icefield

5 06 2012

This publication represents the most recent research output of an ongoing international collaboration between researchers from New Zealand, USA and Norway, which has been using the records of past glacier fluctuations in the Southern Alps to understand climate change from the Last Glacial Maximum to present (c.30ka-present). The first step of this project was to produce detailed maps of the local glacial geomorphology in order to document the past size and extent of glaciers in the Southern Alps (e.g. Barrell et al. 2011). The second step was to date these landforms in order to correlate deposits between glacier catchments and also to be able to compare results to other well-dated climate records of similar age (e.g. Schaefer et al., 2006; 2009; Kaplan et al., 2010; Putnam et al., 2010). This paper by Golledge et al. represents the third phase of this project, which utilises computer models to numerically simulate the past glaciers, from which, quantified estimates of past climate can be derived.

To do this, Golledge et al. use a complex numerical glacier model (called ‘The Parallel Ice Sheet Model’ or ‘PISM’, for short), which has been jointly developed by researchers at the University of Alaska, Fairbanks and Potsdam Institute for Climate Impact Research. This model uses physical equations that mathematically describe the different components of ice flow. Golledge et al. perform a series of experiments, whereby they manipulate the different glaciological parameters of the model (e.g. ice deformation and flow, internal and basal stresses) against the geomorphological evidence, in order to determine the most appropriate values. Once this is completed, the climatic variables (temperature and precipitation) are manipulated and compared to the geomorphological evidence, in order to provide an envelope of the likely climate change that caused glacier advance to the mapped and dated limits.


Figure 1: Modelled ice extent using temperature depression of 6degC (top) and 9degC (bottom) from the present day (Source: Golledge et al., 2012).

Figure 1 shows that the initial, low-resolution (2km grid square) model runs showed that a temperature decline of 6-7 degrees produced the best match between modelled ice extent (grey) and ice extent identified through geomorphological mapping (black line). The mis-matches at the margins of the various outlet glaciers are likely to be due to the coarse resolution of the model, where the outlet valleys are narrower than the resolution of the model run (2km). Narrowing down the likely temperature change at the time of this glaciation (e.g. 6-7degrees lower than at present), allowed Golledge et al. to maximise their computational efficiency in simulating the past icefield at much higher spatial resolution (500m). At this scale, they find that the optimum climatic conditions (i.e. those where the modelled ice best fits the geomorphological evidence) are a 6 – 6.5degC cooling, accompanied by a precipitation change of 0-25%, relative to the present day.

Comparison of these values to those derived from other palaeoclimatic proxies allows Golledge et al. to make inferences about the past climatic regime of the region. The temperature change compares favourably with those from other proxies. For example, temperature estimates from bacteria accumulations in LGM lake deposits suggested LGM cooling of 5.6degC in the Southern Alps (Zink et al., 2010). Meanwhile, regional numerical climate models suggest a cooling of 7.5degC (Drost et al., 2007). In comparing their results to past sea surface temperatures Golledge et al. find that their results are more aligned with those from sub-Antarctic influenced water masses (6.8degC) than sub-tropical (4deg C). They suggest that the local climate at this time was therefore still significantly influenced by southerly winds. Regarding past precipitation, Golledge et al.’s finding of potentially drier conditions at the LGM in New Zealand are consistent with other studies (e.g. Whittaker et al., 2011) that have suggested this was caused by a reduction in strength of the prevailing westerly winds across New Zealand, during the LGM.

In some ways this paper represents the culmination of many years work, from the initial mapping stages, through the development of high-resolution chronologies and now to a palaeoclimatic context. The significance of the terrific geomorphological records of past glaciations in New Zealand’s Southern Alps has long been recognised – as evident in the numerous and influential studies (e.g. Porter, 1975). Understanding regional climatic response to global events such as the Last Glacial Maximum allows us to better understand processes driving past climatic change and New Zealand represent one of the few landmasses in the Southern Hemisphere where such records exist. Future work will focus on improving the chronological constraint of glacier fluctuations in the region and using high-resolution, catchment-scale glacier models to better understand the responses of different glacier types to climatic forcing.