Never-Ending Climate Change

Climate change is a naturally occurring process that has been going on throughout the Earth’s history. In order to predict future climate change, we need to study these changes in the past. Recent variations, geologically speaking, with major Earth boundary conditions similar to those of today, are particularly relevant.

Geological knowledge demonstrates that the climate has been changing throughout the Earth’s history. The temperature varies continuously and at all time scales, and has exhibited many rapid and large-amplitude changes over geologic time. It is also apparent that recent temperature changes are minor compared with those that have taken place in the more distant past. This is all clearly evident from the accompanying graphs on the following pages.

Variations on all scales

On a Phanerozoic time scale (hundreds of millions of years), it is evident that the Earth has gone through both warm and cold periods. The cold periods or ice ages, have occurred at many intervals in the Earth history, such as in the Late Precambrian, in Ordovician – Silurian times, in Carboniferous – Permian times and in Late Cenozoic time.

On a Cenozoic time scale (tens of millions of years), global temperature has dropped much more than the Pleistocene glacial – interglacial contrast of some 6-8°C. The Paleogene was initiated by a gradual warming, but an abrupt global warming event – the Paleocene-Eocene Thermal Maximum (PETM) – marked a rapid and significant climatic perturbation. Global temperatures rose about 5°C in both the surface and the deep ocean. Global warming continued after the PETM for another 5 million years, reaching climate maximum in the early Eocene about 50 million years ago. This super greenhouse mode was followed by global cooling, and around 15 million years ago, ice sheets in high northern latitudes began to form. The plate tectonic setting that the world had now entered into resulted in thermal isolation of both the North and South Pole regions, a situation favouring bipolar glaciations. This is, as far as we know, unprecedented in Earth history.

On a Pleistocene time scale (hundreds of thousands of years), temperatures have varied between highs and lows in a quite regular pattern. In this period, the interglacial stages were much more short-lived than the glacial stages. The tendency has been for the warming ending glacial stages to be rapid, whereas cooling into glaciations took much longer.

On a Late Pleistocene time scale (tens of thousands of years), it becomes clear that the temperature fluctuates rapidly. Changes of several degrees Celsius may occur in a few decades only, on local and regional scales.

Evidence of climate change is, however, not restricted to geological data. Paleobiological data and historical documents of many types show how the climate has varied through the last few thousand years. For example, both historical and archaeological data demonstrate that conditions were favourable for Norse people to settle in West Greenland and sail across to North America about a thousand years ago. This happened during the Medieval Warm Period (AD 800-1300). Conditions in Europe during the Little Ice age (AD 1400-1850) were much more unfavourable.

The two extreme end-member Earth climate states may be characterized as the «greenhouse» and «icehouse» modes. Whether the Earth is capable of being in one or the other of these climate modes is determined by its major boundary conditions. These are primarily the distribution between land and ocean, the location of continents with respect to latitude, their topography, ocean gateways and circulation, and greenhouse gases.

Why study paleoclimate?

Past climate variations has to be understood properly in order to reliably predict the future climate. This is because paleoclimatic studies provide direct evidence on how the natural climate system operates over a variety of time scales.

Climate variations of the recent past are especially relevant for the near future, as most major Earth boundary conditions, such as land – ocean configurations, ocean gateways and currents, have been the same as today. With these settings being the same, the effects of other climate forcing factors, for instance radiation or greenhouse gases, can be compared between different periods of the past and the present.

Paleoclimate studies contribute to our understanding of future climate change by providing real-world scenarios of past environments that are used in testing and improvement of climate models.

Illustration: Eiliv Larsen/Astrid Lyså

The recent past: Variability in the icehouse mode

During the last few million years, the Earth’s tectonic boundary conditions have been favourable for glaciations. This is related to the closing of the sea-way between South and North America forming the Panama strip of land which took place between approximately 13 and 3 million years ago. The result was reduced mixing between the Pacific and North Atlantic oceans, and establishment of a strong North Atlantic thermohaline circulation giving increased precipitation at high latitudes.

The Milankovitch theory proposes that under such conditions, ice ages are triggered by minimum summer solar energy received by northern latitudes, causing winter snow to survive the next summer and thereby start building northern hemisphere ice sheets. This is amplified by increased albedo (reflectivity) as areas with perennial snow, sea-ice and glaciers become larger, and reduction in atmospheric CO2 content due to increased dissolution as world oceans cooled. Conversely, when solar energy is above a critical value, more snow melts during the summer than accumulated the previous winter, and the process is reversed. Paleoclimatic records document glacial – interglacial cycles at least 3 million years back in time in oceanic sediments, and for more than six hundred thousand years in ice cores.

The amplitude of change is best characterized by the growth and decay of continental size ice sheets and accompanying global sea-level changes. The former is illustrated in the figure showing ice-sheet extent at the last glacial maximum. Only in the Pleistocene (the last 1.8 million years), there have been more than 20 glacial and 20 interglacial stages. Sea level variations have been in the order of minus 130 to plus 5m relative to present-day sea level.

The five youngest interglacial stages have apparently been the warmest (compare shaded areas in the figure). During the last interglacial (ca. 116,000-130,000 years ago), the temperatures at high northern and southern latitudes may have been up to 5°C higher than present. The higher temperatures caused reduced ice-cover in Greenland and other northern localities, and probably also in Antarctica, resulting in a global sea level some 5m higher than present. At about 116,000 years ago, high northern latitude solar energy was at a minimum. This initiated the last glacial stage with glaciers starting to expand. The glacial stage culminated with the largest ice volumes globally around 21,000 years ago. At that time, much of North America, Canada, Eurasia and the Barents Sea were covered by glaciers, and sea ice expanded southwards. At the peak of the last glaciation, air temperatures may have been as much as 20 and 8-9°C lower than today in Greenland and Antarctica, respectively.

World oceans were generally cooler compared with today, but tropical oceans were only moderately cooler or were the same as today. Consequently, present day climates also existed during full glacial time, but the arctic climate zones expanded at the expense of equatorial climate zones.

Several rapid, high-amplitude temperature changes occurred throughout the last glaciation. Recent research shows that these were not caused by global changes in the energy balance (i.e. solar energy input as explained by the Milankovitch cycles), but rather by ice sheet instabilities and the release of enormous amounts of fresh-water into the oceans.

The amount of CO2 in the atmosphere was low during the glacial stages and high during the interglacial stages. Carbon dioxide variations thus closely follows temperature variations, but cannot explain the end of glaciations as temperature rise precedes CO2 increase by several hundred years throughout the last 650,000 years.

Earth climate evolution

The upper graph shows generalized Earth climate evolution from late Precambrian time until present. The cold climate periods correspond to ice ages. The evidence is for multiple cycles of glaciation and warming throughout the Earth’s history.

The lower graph shows Paleogene through Neogene deep ocean temperature development based on deep-sea benthic foraminifera 18O isotope records. The compilation was made by Zachos and co-workers in 2001.


Temperature proxies

The upper graph shows variations of deuterium, a proxy for local temperature, and atmospheric concentrations of the greenhouse gas CO2, over the last 650,000 years derived from ice cores from Antarctica and recent atmospheric measurements. Benthic δ18O from marine records, a proxy for global ice volumes, is plotted over the same time period. Shading indicates the last interglacial warm periods. The curves are based on compilation in the last IPCC report (2007) by Jansen and co-workers.

The lower graph shows δ18O from Greenland ice cores plotted over the period from the last interglacial until today, as published by Johnsen and co-workers in 2001. Both δ18O records are proxies for global ice volume fluctuation with downward trends reflecting increasing ice volumes on land.


Holocene glacier variations

Holocene relative degree of glacier variations from Svalbard, Southern Scandinavia and the Alps based on compilation in the last IPCC report (2007) by Jansen and co-workers.

The current interglacial

About 16,000 years ago, a global sea-level rise led to ice-sheet instabilities and rapid deglaciation in the Barents Sea area. The ice-bergs and melt-water entering the sea caused cooling in the Norwegian Sea before the Norwegian current became a dominant climate force 14,500 years ago.

Another cold spell resulting in renewed glacier growth by the Eurasian ice sheet was probably also caused by fresh-water release to the ocean. It occurred between 12,500 and 11,500 years ago, during the Younger Dryas period.

From then on, warming was very rapid. In some regions near the waning ice sheets, rise in air temperatures of 6-8°C within decades occurred. Globally, however, this temperature rise took place over longer time, starting 15,000-16,000 years ago, and ending at full interglacial conditions 5000 years later. A rapid cooling 8200 years ago caused by the final drainage of the glacial Lake Agassiz took place before the northern hemisphere drifted into the Holocene climate optimum that lasted from approximately 8000 to 5000 years ago.

Today, the Earth’s rotational axis is oriented such that the northern hemisphere faces away from the Sun in the winter when the Earth and Sun are closest to each other, and towards the Sun in the summer when the Earth and Sun are farthest from one another. This gives low solar energy in northern latitudes with cool summers and mild winters.

Ten thousand years ago, at the end of the last glaciation, summer solar energy was at a maximum resulting in strong seasonality with warm summers and cold winters. This caused warming in northern latitudes, and mountain glaciers became smaller than today or disappeared completely. Although temperatures during the Holocene thermal optimum was somewhat higher than today, these changes are not necessarily relevant analogues for what might happen in the near future as climate forcing was different. Today’s glaciers are favoured by present low summer solar energy, suggesting that their observed retreat is to be explained by different factors than those that caused glacier retreat in the mid Holocene. The cooling that ended the thermal optimum had a generally gradual trend interrupted by a number of high frequency climate shifts. It has been speculated that these shows system instabilities when climate is shifting.

Climate drivers – during the Pleistocene ice age

Much of the higher frequency global climate changes (10,000-100,000 years) are triggered by periodic changes in parameters of the Earth orbit around the Sun. The cyclic variation of the astronomical parameters was calculated by the Serbian astronomer Milutin Milankovitch at the beginning of the 20th century.

The eccentricity of the Earth’s orbit around the Sun varies from being virtually circular to slightly elliptical with a cycle of 100,000 and 400,000 years.

The inclination of the Earth’s axis varies between 21.5° and 24.5° with a periodicity of 41,000 years.

The precession causes the Earth’s axis to wobble like a spinning top slowing down. A full revolution takes 23,000 and 19,000 years.

Eccentricity modulates the amplitude of the precession and thus affects the annual and seasonal energy budget, whereas changes in inclination of the axis affect the latitudinal energy distribution. The cycles of these parameters are stable for tens of millions of years, and thus is a predictable climate forcing factor on long time scales. For instance will the next glaciations be initiated in 30,000 years from now according to these parameters, and given natural greenhouse gas variation.


First mild, then cold

Medieval times are considered as having been mild compared with the present in the northern hemisphere. New data, although still with uncertainties, suggests that this period was only slightly milder than the 1961-1990 mean, but probably not as warm as the last twenty years. Nevertheless, this period was milder in the northern hemisphere than periods preceding and following. There is not enough data at present to evaluate if the medieval warmth was a global phenomenon.

In the northern hemisphere cooling took place after the medieval warmth culminating in the 17th to the 19thcentury. In many areas this or these periods have been termed the «Little Ice Age.» This is because mountain glaciers expanded well beyond their present limits, and probably to maximum extents for the entire Holocene, although continuous glacial records are sparse. Short-lived and asynchronous mountain glaciations between different regions cannot be explained by long-term solar energy changes, but is rather a result of regionally complex interactions between precipitation, temperature and prevailing wind patterns. A tendency to anti-phasing in glacier chronology between the Alps and southern Scandinavia is attributed to different influence of the North Atlantic Oscillation in the two areas.

Making better predictions

After the Little Ice Age, climate warmed to present values. The pressing issue is if this warming is fully or partly human induced as the general trend coincides with the time after the industrial revolution. Without the type of long climate records provided by the numerous paleoclimate investigations, there would be nothing to measure present climate change trends against.

These kinds of investigations therefore provide an essential reference frame for understanding and explaining climate change. With this at hand, we can say that in a geological perspective, some changes or factors related to the present are unusual. Firstly, the present level of CO2 is about 30% higher than for any interglacial in the past 650,000 years, and is probably the highest for the entire Pleistocene. Secondly, present rate of global temperature increase may surpass any past global rates. The global climate shifts of 5-6°C between glacial and interglacial conditions took place in the order of 5000 years.

The short time period for observations of modern warming calls for caution, but we note that global temperature increase during the last century was 0.7°C, almost an order of magnitude faster than the glacial – interglacial transition rate. Absolute temperatures have been much higher in the distant past than today, but under different climate boundary conditions. Thus, these are not analogues to present climate, but adds to our understanding of the complex climate system, and ultimately to better predictions of future climates.

The last glacial maximum – mapping vs. modelling

During the last glacial stage, ice sheet growth and decay caused a series of glacial sediments to be deposited in mainland NW Russia. The area went through a number of large and short-lived environmental changes including sea-level variations, formation and drainage of huge ice dammed lakes, fluvial sedimentation and incision, glacial erosion and deposition, and strong aeolian activity in a barren landscape. The sedimentary records left behind have for example been used to map and date glaciations affecting the area.

At the maximum of the last glaciation (shown by a thick white line), the Eurasian ice sheet covered the area from Great Britain in the west to NW Russia in the east, and northwards to the continental shelf break north of Svalbard. The ice sheet behaved very differently regionally. For instance, the maximum position was reached much earlier in the west (at about 27,000-26,000 years ago) compared with the east (at about 17,000-16,000 years ago). This has mainly been ascribed to the sources of precipitation being western, and that the continental shelf break limited further ice growth westwards. Yet, the maximum extent represents response to the same climatic forcing.

Modelling of the ice sheet (shown as contours superimposed on the geographical map) is in fairly good agreement with the mapped ice marginal configuration, albeit a single dome is probably not correct.

During the last glacial maximum, global sea level was about 120 m lower than the present corresponding to the amount of water locked in the huge continental ice. The water partly drained back into the sea as massive outbursts with significant feed-back on climate through impacts on ocean circulation.


Maximum extent of the Eurasian ice sheet as mapped (white lines) and modelled (contour lines) after Svendsen and co-workers (2004). The stratigraphic column represents deposition over the last interglacial – glacial cycle in the Arkhangelsk region, NW Russia (Simplified after Larsen and co-workers 2006).

Skrevet av Astrid Lyså, Eiliv Larsen