Natural variability in a constant climate: changing historic perceptions

A paper by Roe & O’Neal (2008) looks at the use of glacier modelling and numerical weather modelling to discriminate between glacier mass balance variations due to natural climate variability or due to global large-scale climate change. Using the Mount Baker glaciers in Washington State, USA, as a basis for the modelling; natural variability was shown to be able to produce 2 to 3 kilometre changes in glacial length over decadal and centennial timescales. As a result, the 1.3 to 2.5 km changes in length on Mount Baker during the Little Ice Age (LIA) can be attributed to this natural variability instead, with no need for global climate change. Instead, they could be the product of yearly weather variation.

The paper notes that all of the processes which contribute to mass accumulation and ablation are ultimately controlled by climate and that large-scale climate changes will often drive glacier variations. However, changing climate doesn’t have to always be the cause of these glacier variations. The authors look at how the statistical distribution of atmospheric variables in a constant climate do not change over time, and how this also means that variability is also an intrinsic part of a constant climate. Historical records and reconstructions of glacial response times show that glaciers are able to reflect this variability. And whilst this variability in glacier response time may vary from years to centuries, the paper asks the key question of whether these past movements are statistically different from the normal response of a glacier to the variability shown to exist in a constant climate.

Whilst the paper used historical climate data from sites situated over Mount Baker in the Cascade Range of western Washington, USA, to show glacial response through the use of a linear model; its intention was not to simulate any one glacier. Rather, the observations were used to show the much wider effects of climate variations in a stationary climate on glacier length.

Figure 1. Glaciers of Mt. Baker (USGS, 1999)

The authors recognise the similar approach taken to that of Reichart et al. (2002), who ultimately drew very similar conclusions about the preindustrial fluctuations of LIA glaciers being explained as within the internal variability of the climate system. The Reichart et al. (2002) paper used a global climate model and a dynamic glacier model for two European glaciers to show this. An important point from this study is that whilst LIA glacier advances did not exceed natural variability, the present retreat since then certainly does. Therefore, whilst climate change isn’t needed to explain LIA variability, it is needed for changes today and in the foreseeable future. Furthermore, whilst studies of this type are useful to change our knowledge of past glacier sensitivity, they do not show the kind of glacial retreat that we can see today as a result of anthropogenic activity.

It is quite common in the literature to find decadal climate variability given as the reason for glacier variability on these timescales. This includes the Pacific Decadal Oscillation (PDO) which is the main orbital forcing factor on sea-surface temperatures in the North Pacific and on climate patterns in the Pacific northwest. What this study and others (Huybers and Roe, 2009) showed, was that there was actually little or no significant interannual memory for the atmospheric variables that control glacier variability and so they therefore don't fully explain the observed variance. Indeed, the authors state that the appearance of decadal variability in time series of the PDO is often artificially exaggerated by the application of a multiple-year running mean through the data.

At one of the monitoring sites used in this study, a test for autoregression was applied to the long term record of climate, and showed that there was no statistically significant interannual relation for either precipitation or melt-season temperature. Put simply this shows that one year had no relation to the next. The paper also criticises the often-used technique of applying a five-year running average through climate data, which may give an artificial view of decadal climatic periods. They also accept that whilst there are chance intervals of high or low conditions over several years, the use of a running average highly exaggerates this, rather than accepting the existence of high year to year variability. To reiterate the point here, there was little to no observed interannual relationship and thus no evidence for interannually linked climate regimes.

As can be seen in the wider glaciological literature as well as for Mount Baker, much of the pre-anthropogenic glacial history has been explained as the result of decadal climate regimes. However, in the findings of this study, it is shown to be the intrinsic inertia or memory found for each glacier and not the climate system which drives previous long term variations. The simple model used in this study showed significant centennial variability, with an amplitude of 2-3 km for glacier length. As the authors state very clearly:

“Thus there is centennial, and even millennial variability in the spectrum, all fundamentally driven by the simple integrative physics of a process with a perhaps-surprisingly short timescale, and forced by simple stochastic year-to year variations in climate.”

Evidently, variability on such a time-scale can be used to reinterpret past glacial histories and this is attempted in the paper. The authors state that from the instrumental observations and the model parameters used to show natural variability:

“...the 1.3- to 2.5-km length fluctuations on Mount Baker attributed to the LIA can be accounted for by the model without recourse to changes in climate.”

What is stated here is that the kilometre-scale length variations of glacier terminus don’t need large changes to temperature and precipitation to trigger them. Instead the natural interannual variability of a constant climate will do this.

What was also noted, was the importance of measuring small-scale climate forcing patterns, to fully understand individual glacier response. The model produced results which suggest that random climatic fluctuations over the past 1000 years may have been enough to produce large changes in glacier length. Indeed, many moraines may be products of much earlier advances and not necessarily synchronous with each other, nor part of a global pattern. The paper notes that the current world-wide retreat, as opposed to the past regional and local variance observed in this study, is a strong indicator of global climate change.

In summary, the suggestion that long-term glacier variability can occur in a constant climate was explored. The authors found that it would be impossible to rule out the effect of variations observed in a stationary climate on past glacier length. The obvious effects on how we understand past glacier changes is looked at for the LIA although not developed further. Overall, the paper offers a very exciting look at how we can see glacier dynamics in many different ways. But whilst it may dispute past climate forcing on glaciers, it reinforces the point that global retreats in glacial length seen today are indisputably as a result of anthropogenic climate change.

Figure 2. The peak of Mt. Baker

Blog poster

The below academic poster was produced and presented to show some of the ideas looked at in this blog already, as well as other papers that have yet to be covered. Either way, it provides a nice overview of the kind of things this blog is covering! (click on image for full size version)

What's in the news?

From the archives of The Guardian, an online article from May 2005 explains a novel response of one Swiss ski resort to the melting of the glacier above their village that is perhaps worth mentioning. The village of Andermatt is located underneath the Gurschen glacier, which experiences considerable summer melting.

The solution: a 26,900 square foot plastic sheet to put over their ski slope and hopefully stop it from melting away over the summer. Whilst a short term solution to skiing problems, the solution seems far detached from dealing with the causes of a warming climate. The problem isn’t just limited to this location either. As the article states:

“Experts at Zurich University recently estimated that Swiss glaciers had lost about a fifth of their surface area in the past 15 years. In that time, the 2,961-metre high Gurschen glacier has sunk by 20 metres.”

This solution does have its more immediate drawbacks as well. The sheet isn't cheap at a cost of £42,000, however this is no doubt outweighed by the monetary advantages of having a ski resort. However, despite these criticisms, it appears that the sheets are successful; having been tested in Austria the previous year.

Read the full article here: The Guardian

Climate change - not the only influence on glacial dynamics...

When looking for how past and present changes in glacier dynamics might be able to inform us of future behaviour, it is important to understand the processes that affect the movement of individual glacial terminus. Whilst the most obvious and widely-attributed influence on global glacial retreat is an increase in air temperature, a study by Scherler et al. (2011) suggests that this is not the only influence on the response of glaciers to climate change. The authors look at the effect of debris cover on terminus dynamics of a number of Himalayan glaciers - which had not yet been established in the region - and look at the regional differences of glaciers at the mountain-belt scale.

As was mentioned before in this blog, glacial meltwaters and snow are an important resource for people around the world. This is particularly true in Central and South Asia, where as the paper states:

“Snow and glacial meltwaters make an important contribution to the drinking water, agriculture, and hydropower supply of densely populated regions...”

This is a very immediate concern for the people of this region, as whilst mountain-river discharge will increase in the short term (perhaps leading to flooding in low-lying areas), in the long term the discharge will decrease; which may well have widespread humanitarian impacts as a result of water scarcity in the region.

Several glaciers in the central Himalaya were found to have stagnant reaches up to several kilometres. Despite evidence of glacial shrinking, such as increased meltwater ponds and surface lowering, the fronts of these studied glaciers have remained stable. All were debris-covered, and thus as the authors of the study note; debris cover can modify a glacier’s response to climate change. However, despite this growing body of evidence, no study to date has looked into the impact that debris cover can have regionally on glaciers.

A number of aims of the study can be identified. The first was to determine if glaciers in the greater Himalaya region display evidence of spatial patterns relating to altered mass balance and frontal dynamics and whether these patterns are related to climatic variations or other factors. Secondly, the study aimed to see if spatial variation of debris covered glaciers can be directly linked to regional differences in frontal dynamics.

The study looked at 286 mountain glaciers and indentified 6 regions that differ in climate and topography. For the climate forcing hypothesis presented later in this post, it is important to note that as the study moves over from the western most sites (Karakoram and Western Himalaya) towards the central Himalaya, there is a notable decrease in the influence of mid-latitude westerlies and the increasing influence of the Indian monsoon.

Across all moraines, rates of frontal change were observed between -80 and +40 m yr^-1. In Karakoram however, 58% were stable or slowly advancing with a mean rate of approximately +8 m yr^-1. This greatly contrasted with all of the other regions. The study thus observed the highest concentration of retreating glaciers in the areas where debris-covered glaciers were lowest.

The simplest models of glacier dynamics denote that the time scale of a glacier’s response to climate change is inversely proportionate to the surface slope, as well as being affected by local climate and glacier size. However, these factors do not alone explain the differences observed between debris covered and debris free glacial fronts. The summary at to this point is made clear in the paper:

“...widespread debris cover on many Himalayan glaciers reduces their retreat rates, which are therefore unsuitable as indicators of recent climate change.”

What is made clear here is that whilst the differing movements of these glaciers are indicative of varying debris cover and topography (increased hillslope angle leading to increased flux of rocky debris), they do not represent the effects of differing climate.

Figure from Scherler et al. (2011)

However, the paper goes on to suggest that further regional climatic differences can be seen to have an effect on other glaciers in the region. It is suggested that the approximate 50% stable or advancing glaciers in the westerly Karakoram are not related to stagnant terminus regions (as a result of topographic factors) but are a consequence of different mass-balance regimes associated with their climatic setting. The paper proposes that historical changes in the westerly-derived winter precipitation may cause the shifts towards positive mass balances. Three points of evidence are given to support this theory:

1) That the westerly jet stream, which provides the highest moisture transport during the winter, has strengthened and shifted to lower elevations. This therefore increases the potential for snow and ice accumulation in the study region.

2) Karakoram tree ring records support the theory of an increase in 20^th century winter precipitation.

3) Summer temperatures in the areas showing less frontal retreat have decreased by a small amount, linked to higher precipitation and increased cloudiness (reduced insolation).

The importance of understanding glacial sensitivities across many different regions is very important. In all but one of the six regions studied in this paper, most of the stored ice is found in glaciers with more than 20% debris cover. Thus, knowing how debris cover affects the response of a glacier to climate change is significant for our predictions of future water resources in the region. What we currently know is that debris cover does slow melt rates and therefore the response of a glacier to temperature rises is reduced. Thick debris cover can also have other limiting effects, including containing the effects of decadal to centennial variations in solar radiation and the effect of anthropogenic forcing from atmospheric dust and soot deposition so that only minor changes in mass balance are observed.

The wider applicability of this study is promising, as debris-covered glaciers are common in other mountain ranges globally. To give truly representative estimates of future water availability and sea level rise, then the models that calculate mass balance estimates must include debris cover. At present they do not account for this.  

What's in the news?

An article regarding potentially ground-breaking changes in the understanding of how the East Antarctic ice sheet moves and expands appeared on the BBC News website last week (3^rd March). The news item draws from a paper by Robin et al. (2011) published in Science magazine a few weeks earlier.

The data collected by airborne radar showed the layering of ice, through the sheet to the bedrock. What this revealed was that liquid water – which had been known to exist under the ice sheet – was freezing in great amounts to the bottom of the East Antarctic ice sheet. Whilst this process was not fully understood before now, it was assumed that the freely moving water under the ice sheet, eventually fed back into the southern ocean. This previous focus for the processes involved with subglacial water is exemplified by a relatively recent paper by Wingham et al. (2006) which looks at periodic mass discharge of subglacial lakes as the primary outlet for basal water transfer. As one of the authors of the new paper stated for BBC News:

"...it was demonstrated this water could move, it could slosh around; but I think we still had this idea that it just spilled into the ocean... Well, now we can show these hydrologic systems are modifying the fundamental stratigraphy of the ice sheet.”

Evidently, the mass discharge of this subglacial water into the ocean proved to be incorrect, with the authors of the new paper noting that in some of the places they monitored; the newly formed layer accounted for over half of the entire ice column. With this new understanding of how the East Antarctic ice sheet behaves, it will be possible to revise estimates of how the glacial environment will respond to climate change. The paper states that at Dome A, this basal ice forms 24% of the ice sheet base. With the knowledge that in some areas basal ice formation accumulates at a faster rate than snow deposition, surely greater care will have to be taken in measuring glacial expansion.

This new discovery could also have impacts for paleoclimatic research using stable isotope analysis to date accumulated ice layers; which often require deep cores into the ice sheet. As the article states, the impacts are both good and bad. On the one hand, old ice could be pushed upwards by the new formation and make the archives of paleoclimatic data much more accessible. On the other hand, there may be noticeable damage inflicted to this old ice, including: “melting deformation and destruction of ice sheet records.”

This new information clearly has impacts for the way we understand Antarctic ice sheet dynamics; including both how the ice may have expanded in the past, but perhaps more importantly, how we understand and predict future changes. The size of both the Antarctic and Greenland ice sheets can’t be forgotten, as well as the potential effects either could have on global sea level rises. It is therefore important to fully understand the processes that underlay such large glacial environments.

What do glacial mass balance records tell us?

Records of glacial mass balance provide us with a clear view of how glaciers in most regions have changed over time. Whilst these records predominantly exist in the northern hemisphere, there also exist good studies elsewhere as well. By knowing how the mass balance has changed, climate change influences can be inferred as well as other factors such as sea level rise correlated. 

As Zemp et al. (2009) discuss in their review of the WGMS programme, from looking at the data of 30 continuously recorded ‘reference’ glaciers there has been an evident trend of increasing glacial mass loss over the past several decades:

“...continuous measurements since 1976 show an (arithmetic) average annual mass loss of 0.58mw.e. for the decade 1996–2005, which is more than twice the loss rate for the period 1986–95 (0.25mw.e.), and more than four times the rate for the period 1976–85 (0.14mw.e.).”

Not only that, but this trend is clearly set to continue; with the paper stating that a preliminary value of -1.30 m w.e for the year 2006 is recorded (which therefore breaks the record for highest annual mass loss yet). Whilst this data provides clear evidence, starker evidence is expressed in the paper:

“The vast ice loss since the mid-20th century has already led to the disintegration of many glaciers within the observation network, including Lower Curties and Columbia 2057 (US), Chacaltaya (BO), Care`ser (IT), Lewis (KE) and Urumqihe (CN)...”

Figure 1. Global glacier mass changes from 1945 to 2006. The cumulative mean specific mass balance (left y axis) of the reference glaciers and of four different sampling/averaging approaches (see text) are shown together with the number of available observations (right y axis) from reference (black) and other (grey) glaciers. Source: data from the WGMS.

The paper also explains the main processes behind this drastic reduction in glacial mass. Whilst changes in mass balance provide a climatic signal that is made up of several different sources, including solar radiation, air temperature, precipitation, wind etc... air temperature thereby has an integral part to play as it is related to the radiation balance, turbulent heat exchange and solid/liquid precipitation ratio (Zemp et al. 2009). The point of this is to show the clear effect that a change in climatic conditions and temperature has on a glacier, in most cases.

To try and make this point clearer, for a temperate glacier; initial changes in climatic conditions would cause a change in mass balance, followed by a return towards zero mass balance change values. This return to stable glacier mass conditions happens as a result of the glacier surface area adapting to the new climate (shrinking with warmth and vice-versa). However, when we see in these records that the negative mass balance is increasing whilst the surface area decreases it is evident that climatic forcing continues to affect glaciers and perhaps even shows the effects of positive feedback; such as changes in surface albedo, more intense turbulent fluxes from greater rock outcrops, or changes in the extent of the equilibrium line altitude (ELA).

Additionally to the WGMS report, Norway provides the only long term, nationally funded records, of changes in the mass balance of Norwegian glaciers and as such provides a useful, high resolution record, of growth and retreat from past to present. The vast majority of mass balance records in Norway go back to the early 1960s and are largely uninterrupted. A study by Nesje et al. (2008) shows that the RegClim climate scenario study estimated a rise of 2.3°C in mean summer temperatures and an increase of precipitation by 16% during accumulation season in Norway up to 2100. This rise in summer temperatures and increase in winter precipitation is in turn expected to create melting conditions of ~140±30 m w.e by 2100 and thus a net loss of glacier ice.

The subsequent figure of 98% reduction in the number of Norwegian glaciers by 2100 makes for hard reading!

Figure 2. Briksdalsbreen, a western outlet glacier from Jostedalsbreen, has retreated 300 m between 1997, when it was located at the outlet of the lake, and July 2006. Photo: Atle Nesje. (Nesje et al. 2008)

Introduction to the blog

Before going into the more specific issues that this blog will cover, it is first important to give a summary of what I hope to achieve! This blog will seek to explore the sensitivity of glaciers globally, to both climate change in the past, as well as to the warming of climate today. Alongside this, by looking at how representative of a changing climate glaciers have been both before and now, we can assess the reliability for looking at future shrinking (or growth). The data on changes in glacial mass balance and other indicators of change will be sourced from a range of academic papers, and analyses of different phenomena interpreted in this blog. As well as the use of academic material, a look at how much of this information feeds through to more mainstream and social media and whether it is always correctly interpreted will be taken.

The significance of knowing how glaciers will behave in the future is important, considering the impact that the loss of these environments could have as both a freshwater resource for many people (particularly in Central Asia) and as an important habitat for many different flora and fauna. Knowing how glaciers in the past expanded and shrunk also helps us to reconstruct past environments. As well as this, by knowing the sensitivity of glacial environments to varying degrees of temperature change, we can see how vulnerable they are to our present anthropogenic warming.

A useful resource to initially explore is the University of Zurich’s World Glacier Monitoring Service (WGMS) report on Global Glacier Changes which provides an accessible set of data on glacial growth and reduction in the long term. By first seeing how trends in glacial growth/reduction have been up to the present day, an appreciation of how significant current reductions are can be made. This blog will aim to move on from this face-value data; and show many important interpretations and academic insights about glacial change!

Figure 1. Global measure of increasing negative mean cumulative specific mass balance as millimetres water equivalent (mm w.e.) over the last few decades. (Zemp et al. 2008)