By Andy May

Glacier length changes through time, they advance when the local climate around them is colder and retreat when it is warmer (Bray, 1968). Over century and greater time scales glacier length is considered a highly reliable indicator of both regional and worldwide warming trends according to Olga Solomina, Johannes Oerlemans, and the IPCC (Solomina et al., 2008), (Oerlemans, 2005) & (IPCC, 2001, pp. 127-130). While studying glacier lengths can illuminate long-term warming or cooling trends in glaciated areas is true, the idea that they can reveal hemisphere-wide or global climatic trends is somewhat speculative.

Advancing and retreating glaciers leave evidence of their fluctuations in length in glacial till deposits called moraines. Glacier moraines are easily identified and are distinct from other sediments and sedimentary rocks because they contain angular boulders, and they are unsorted and unstratified. Olga Solomina and colleagues in a 2015 review article, note:

“Studies of Holocene glacial geomorphic and sedimentological records provide the most direct means of determining the extent and timing of glacier oscillations. Until recently it has been difficult to define the ages of moraines in many regions because of the lack of appropriate dating techniques. Radiocarbon has been the most widely used and in some cases optically stimulated luminescence (OSL) dating has been implemented, but in most cases these can only be utilized to provide maximum and/or minimum ages on moraines by dating organic-rich deposits that are buried beneath moraines/tills, beyond the glacial limit (maximum ages), on top of moraines, or within the glacial limit (minimum ages). The development of terrestrial cosmogenic nuclide (TCN) dating, however, has provided a direct method of dating moraines and has led to a plethora of studies that are shedding new light on the nature of Holocene glacier fluctuations.” (Solomina et al., 2015)

Dating Glacial Advances

The TCN (terrestrial cosmogenic nuclide) dating technique (Larsen et al., 2021) is uniquely suited to dating the maximum extent of glacial advance, prior to a retreat. It is a geochronological method used to determine the exposure age of materials on Earth’s surface, such as rocks, sediments, or landforms. It works by measuring the concentration of rare isotopes (or more accurately nuclides, including isotopes of beryllium, chlorine and carbon) produced by interactions with cosmic rays. The nuclides accumulate over time and provide a “clock” for how long the material has been exposed to cosmic radiation. The technique is particularly valuable for dating Quaternary events from a few hundred years to several million years before present, depending on the nuclide and site conditions.

Target materials for TCN dating are large quartz rich boulders that have been ripped off bedrock by the glacier and lie exposed on a moraine crest. These would be very angular boulders that were not previously exposed until deposited in or on the glacial till. The TCN derived age identifies when the boulder was deposited and became stabilized on the top of the till. Factors that can interfere with accurate dating are significant erosion or long-term burial under ice and snow. Moraines are not a stable geological feature once deposited, especially if they are cored with ice. Boulders may shift with time and may contain nuclides from a previous exposure. Careful sampling and proper analysis of multiple boulders per site can usually detect, and sometimes correct, for these problems (Larsen et al., 2021).

Glacier advances and retreats are very long-term climate indicators. They are quite sensitive to small local changes in average temperature and can be dated accurately. They are most useful in detecting when a glacier changes from a long-term advance to a long-term retreat leaving a “terminal moraine.” A maximum retreat is harder to see, since subsequent advances often disrupt the terminal moraine of a retreat and boulders may contain nuclides accumulated during previous exposures (Larsen et al., 2021).

An Anthropogenic Warming Indicator?

Solomina, et al. believe that the current rate of glacier retreat is unusual and an indicator of anthropogenic warming. This is quite speculative due to the very short period of possible anthropogenic warming, roughly the past 70 years according to the IPCC (IPCC, 2021, p. 117). Since the world cooled from 1950 to ~1975, the period of warming is actually shorter, more like 50 years, and in the middle of this period, from 1998 to roughly 2013, there was another period of cooling (or at least a “pause” in warming) which casts further doubt on the hypothesis that humans have significantly affected climate with their greenhouse gas emissions. For a summary of a discussion of the so-called “hiatus in warming” by some prominent climate scientists see here.

Solomina et al. and the IPCC (IPCC, 2007b, p. 436) believe that the recent warming and the associated nearly global glacial retreat cannot be attributed to the same orbital causes as those present during the Holocene Climatic Optimum (or “HCO”, see here, figure 4), so they must be due to human greenhouse emissions. It is true that Earth’s orbital condition is different than during the HCO, but the choice of causes is not binary. The twentieth century contained the Modern Solar Maximum, the longest solar grand maximum in 2,000 (SN-S series) to 8,800 years (SN-L series) according to Usoskin et al. (Usoskin et al., 2007, Tables 2&3). Just because the orbital position has changed since the HCO does not mean that modern warming is due to human activities.

What is climate?

Climate is generally defined as the average or prevailing weather of an area over a long period of time, the minimum period to define climate is normally taken as 30 years, the “area” described by the term is undefined. Thus, to measure a change in climate one must have two non-overlapping periods of greater than 30 years each to compare. Even 30 years may be too short since the very influential AMO weather cycle is 60-70 years long. Other significant long-term weather oscillations are discussed here. Glacier length records are good long-term climate indicators if the length changes are a small fraction of the total average glacier length. Glacier length changes are useful on century to multi-century times scales (Oerlemans, 2012). There are some rare glacial records that are accurate at decadal time scales, but these are usually short records, span only the last one or two millennia, and are concentrated in the Alps and Scandinavia. In other cases, claims of decadal resolution are belied by neighboring inconsistent glacier decadal records (Oerlemans, 2012). Another complication is that glaciers are rarely in equilibrium with their environment and the response time to changes in local climate can be hundreds of years for larger glaciers on gentle slopes (IPCC, 2021, p. 1278) & (Oerlemans, 2005).

How large an area is needed to define a “climate.” This is a tough question, figure 1 suggests that climate changes are not uniform over the whole planet. The middle latitudes of the Northern Hemisphere (NH) march to their own drummer relative to the rest of the world and the same can be said of Antarctica and the Southern Hemispheric mid-latitudes (SH). I have often shown plots of the Rosenthal Makassar Strait and Vinther Greenland Holocene temperature reconstructions and believe that they represent climate changes in their immediate areas. But I doubt that plots of gridded or average hemisphere-wide or global temperature proxies mean very much, the term “climate” just isn’t well defined over such large areas.

Holocene Glacial Advances

Most glaciers have been retreating since the middle 19^th century in the Northern Hemisphere and in some parts of the Southern Hemisphere but are still larger today than they were in the early and middle Holocene. Most glaciers reached their minimum Holocene extent between 8 kyrs and 6 kyrs (that is between 8,000 and 6,000 years ago). Subsequently glaciers expanded and reached their maximum Holocene extent between about 1500AD to 1850AD or so (IPCC, 2021, p. 345) & (Solomina et al., 2015). Importantly, most glaciers all over the world reached their maximum Holocene extent in the Little Ice Age, this includes all over the Northern and Southern Hemisphere and in the tropics. In figure 1 the maximum glacier advances peak between 1000AD and 2000AD for all regions. In Alaska, Greenland, Iceland, Scandinavia, Central Europe, Russia, the tropics, and Antarctica glaciers were smaller than today during the Roman Warm Period, roughly 250BC to 400AD (Solomina et al., 2015, Fig. 2).

Short-term and recent glacial retreats often yield poor results with wide age scatter, so accurately measuring retreat rates from the major Little Ice Age (LIA) glacier advances is problematic and limited to the instrumental era (Oerlemans, 2005) & (Oerlemans, 2012). Glacier retreats are rarely clean; they are characterized by frequent re-advances that contaminate young fresh boulders with disturbed boulders from earlier retreats or advances yielding a wide scatter of dates. Thus, we probably will not know how the current rate of glacial retreat compares to past retreats for several hundred years, if at all.

Comparing Glacier Advances and Retreats to other temperature records

The elements of global and Northern Hemispheric temperature changes during the Holocene are generally agreed by most scientists. There was a very rapid rise in temperature after the Younger Dryas cold period around 9,700BC at the beginning of the Holocene (Walker et al., 2009), the warming peaked during the Holocene Climatic Optimum (HCO) sometime before 3,500BC in the tropics and the Northern and Southern Hemisphere mid-latitudes (see here). It ended earlier in the Arctic and Antarctic, perhaps due to changes in obliquity that decreased insolation at the poles and increased it in the tropics (see here for a discussion).

After the end of the HCO, in the latitudes outside the polar regions around 4000BC, the Mid-Holocene Transition (MHT) period began. This was when the Sahara began to turn into a desert and temperatures declined (except in Antarctica) in a period called the Neoglacial into the Little Ice Age (LIA), the coldest period in the Holocene. Temperatures did not turn around and begin to warm until 1700 to 1850AD. The warming period after 1700AD is usually called the Modern Warm Period. The elements of the Holocene are identified in figure 1, the beginning and end of each period are approximate, since the transitions are gradual and do not occur synchronously across the globe.

Figure 1 plots latitude bounded temperature reconstructions using selected Holocene proxies from Marcott’s (Marcott et al., 2013) global collection. The details of how each reconstruction was made are explained in a series of posts here. I’m not fond of reconstructions like these because they can be misleading. The curves in figure 1 are probably directionally correct, but they are very low resolution and not very accurate, so the temperature anomalies must not be taken literally. They will not show climatic events that are shorter than about 150 years and the temperature accuracy is no better than ±0.5°C. Variability in temperature is higher than shown, thus these proxies cannot be compared to modern instrumental measured temperatures, although they often are, see here for more details.

As discussed in my last post and here it is better to discuss climate in a local context than regionally (as in figure 1) or globally. In the last post I plotted Vinther’s Greenland reconstruction and Rosenthal’s Makassar Strait reconstruction, each represents a relatively small area, is reasonably accurate, and each has a temporal resolution of between 20 and 50 years, much better than the average Holocene temperature proxy resolution of 164 years (Kaufman et al., 2020b). The best way to compare today to the past is within local climate regimes. However, to study global climate changes, one must make large area reconstructions like those plotted in figure 1 and remember that their temporal resolution and accuracy are poor.

As figure 1 makes clear, climate changes, at least as defined by average temperature, differ a lot by latitude. The middle northern latitudes (“NH”, 30N to 60N, black thick line in figure 1) stand out. NH, Antarctica, the Arctic, and the tropics warm faster in the early Holocene and the Southern Hemisphere (SH) warms later. Peak warming occurs late in the NH and SH and early in the Arctic and Antarctic. Temperatures drop early in the Antarctic and recover in the mid-Holocene Transition. During the Neoglacial period, prior to the Little Ice Age, most of the world oscillates around a fairly constant temperature or slightly declines as the NH temperature falls rapidly. After 1000AD temperatures fall in the Arctic, SH, and tropics, but rise in Antarctica. The tropics and the Arctic have two peak temperatures, one early in the Holocene and one during the mid-Holocene transition.

Glacial advances tell the same story. In figure 1, the three rows of numbers between -1.5 and -2 degrees are, top to bottom, the Northern Hemisphere, tropical, and Southern Hemisphere glacial advances as tabulated in Solomina, et al., 2015 in their table 2. I added the advances for each millennium. In Solomina et al.’s table 2 the values summed are the first values only, I ignored the possible duplicates that follow the plus signs.

Scanning the glacial advance totals, we can see that before 7,000BC they are more numerous. Between 7,000BC and 2,000BC there are fewer advances. After 2,000BC, as the Bronze Age collapses into the Greek Dark Age, the number of advances increases until it peaks after 1,000AD. By far, the Little Ice Age (LIA) sees the most glacial advances in the Holocene Epoch.

Discussion and Conclusions

The poor resolution and inaccurate (but directionally correct) latitude slice temperature reconstructions plotted in figure 1 are supported in a qualitative way by Solomina et al.’s glacial advance summary. The Holocene ITCZ shift that marks the beginning of the desertification of the Sahara falls at the second peak of the Antarctic, tropical, and Arctic records, as well as at the Southern Hemisphere peak Holocene temperature. Antarctica does not have a Little Ice Age anomaly, but all the other slices do, although they are not synchronous. The Northern Hemisphere Little Ice Age anomaly dwarfs those seen in the other slices. We read a lot about “Arctic amplification” but it is the Northern Hemisphere mid-latitude temperature record that stands out at this scale.

The Little Ice Age glacial advances from 1,000 to 2,000AD are the maximum seen in all three regions (NH, T, and SH) in the entire Holocene. Given the time frame of glacial advance and retreat moraine preservation and detection, glacial advances cannot be used to support or disprove anthropogenic warming for at least another one-hundred years and probably longer. Figure 1 also suggests that the idea of “global warming,” that is a uniform rise in temperature due to changes in greenhouse gases over the whole globe synchronously, is not supported over the Holocene. Figure 2 shows that “global warming” isn’t even a good description of what is happening today.

Figure 2 is from AR6 (IPCC, 2021, p. 316) and shows the warming globally from 1900-1980 (top map) and 1981-2020 (bottom map). Redish colors indicate warming and blue colors cooling. The x’s indicate grid cells with an insignificant trend. White areas are areas with insufficient data. The bottom line is that the aerial coverage, especially in the Southern Hemisphere is poor and that many areas have cooled over the past century, not warmed. It cannot be said that glacier fluctuations support the idea of recent anthropogenic warming, and there is considerable doubt in the instrumental data as well.

Works Cited

Bray, J. R. (1968). Glaciation and Solar Activity since the Fifth Century BC and the solar cycle. Nature, 220. Retrieved from https://www.nature.com/articles/220672a0

IPCC. (2001). Climate Change 2001: The Scientific Basis [TAR]. New York: University Press. Retrieved from https://www.ipcc.ch/site/assets/uploads/2018/03/WGI_TAR_full_report.pdf

IPCC. (2007b). WG1: Climate Change 2007: The Physical Science Basis (AR4). Cambridge University Press. Retrieved from https://www.ipcc.ch/site/assets/uploads/2018/05/ar4_wg1_full_report-1.pdf

IPCC. (2021). Climate Change 2021: The Physical Science Basis. In V. Masson-Delmotte, P. Zhai, A. Pirani, S. L. Connors, C. Péan, S. Berger, . . . B. Zhou (Ed.)., WG1. Retrieved from https://www.ipcc.ch/report/ar6/wg1/

Kaufman, D., McKay, N., Routson, C., Erb, M., & Dätwyler, C. (2020b). A Global Database of Holocene Paleotemperature Records. Scientific Data, 7(201). https://doi.org/10.1038/s41597-020-0530-7

Larsen, N. K., Søndergaard, A. S., Levy, L. B., Laursen, C. H., Bjørk, A. A., Kjeldsen, K. K., . . . Kjær, K. H. (2021). Cosmogenic nuclide inheritance in Little Ice Age moraines – A case study from Greenland. Quaternary Geochronology, 65. https://doi.org/10.1016/j.quageo.2021.101200

Marcott, S. A., Shakun, J. D., Clark, P. U., & Mix, A. C. (2013, March 8). A Reconstruction of Regional and Global Temperature for the Past 11,300 Years. Science, 1198-1201. Retrieved from https://science.sciencemag.org/CONTENT/339/6124/1198.abstract

Oerlemans, J. (2005). Extracting a Climate Signal from 169 Glacier Records. Science, 308. https://doi.org/10.1126/science.1107046

Oerlemans, J. (2012). Linear Modelling Of Glacier Length Fluctuations. Geografiska Annaler: Series A, Physical Geography. https://doi.org/10.1111/j.1468-0459.2012.00469.x

Solomina, O. N., Bradley, R. S., Hodgson, D. A., Ivy-Ochs, S., Jomelli, V., Mackintosh, A. N., . . . Young, N. E. (2015). Holocene glacier fluctuations. Quaternary Science Reviews, 111. https://doi.org/10.1016/j.quascirev.2014.11.018

Solomina, O., Haeberli, W., Kull, C., & Wiles, G. (2008). Historical and Holocene glacier–climate variations: General concepts and overview. Global and Planetary Change, 60. https://doi.org/10.1016/j.gloplacha.2007.02.001

Tomkins, M. D., Dortch, J. M., Hughes, P. D., Huck, J. J., Pallàs, R., Rodés, Á., . . . Rodríguez-Rodríguez, L. (2021). Moraine crest or slope: An analysis of the effects of boulder position on cosmogenic exposure age. Earth and Planetary Science Letters, 570. https://doi.org/10.1016/j.epsl.2021.117092

Usoskin, I. G., Solanki, S. K., & Kovaltsov, G. A. (2007). Grand minima and maxima of solar activity: new observational constraints. Astronomy & Astrophysics, 471(1), 301-309. https://doi.org/10.1051/0004-6361:20077704

Walker, M., Johnsen, S., Rasmussen, U. O., Popp, T., Steffensen, J.-P., Gibbard, P., . . . Newnham, R. (2009). Formal definition and dating of the GSSP (Global Stratotype Section and Point) for the base of the Holocene using the Greenland NGRIP ice core,and selected auxiliary records. Journal Of Quaternary Science, 24, 3-17. https://doi.org/10.1002/jqs.1227