Correlation of temperature and CO2 for last 420,000 years from Vostok

EC Perspective: Accounting for 800,000 years of climate change

Approximately 800,000 years ago something changed in the Earth’s climate system that led to the climate then following a series of approximately 100,000 year cycles. As a result, eight ice ages, along with the intervening warm periods called interglacials, have been the dominant forces in creating, extinguishing and changing nature and life on Earth.   Along with these 100,000 year cycles there have been shorter warming and cooling cycles each lasting in the order of 1500 years. Some of these short cycles have resulted in increases in Greenland temperatures, of as great as 8–16 °C, occurring within decades or even years.   Understanding what causes these climate cycles is fundamental to understanding the global climate in the past and present. There is evidence that small, predictable changes in the Earth’s orbit about the Sun act as triggers for the glacial and interglacial periods.   By themselves however these are not able to explain all the cycles of global warming and cooling.  A more comprehensive plausible explanation is now emerging that also requires taking into account variations in ice sheet size, snow and ice albedo, atmospheric CO2 concentration, ocean currents, extent and productivity of wetlands, melting permafrost, deep ocean CO2 ventilation and atmospheric dust. The availability of increasingly more detailed and accurate paleoclimate data has made it possible to relate most of the variation in the Earth’s temperature at the end of the last ice age, from 19,000 to 11,000 years ago, to these additional factors.  This account, however, has now raised further questions about the physical and biological mechanisms which caused the warming and cooling cycles. One important and perplexing question relates to the fact that surface temperature and CO2 have remained in a lockstep relationship generally to within centuries, and in Greenland even to within decades.   These questions and the most recent science relating to them is the subject of this article.

Background

For most of the Northern Hemisphere ice ages, from 3 to 0.8 million years ago global ice volume varied predominantly with a 41,000 year cycle corresponding to the Earth’s orbital obliquity. But about 800,000 years ago the ice ages and intervening warm periods began to cycle with a longer period of about 100,000 years. Accounting for the glacial/interglacial cycles has been the focus of research of intensified research in the last few decades as new data about the Earth’s paleoclimate has emerged.

The latest paleoclimate research is beginning to reveal a plausible account of Earth’s cycles of ice ages and intervening warm periods. The paleoclimate record, which is comprised of data from Antarctica and Greenland ice cores, marine sediment cores, cave calcite cores, and other sources, provides information about regional surface temperature, global atmospheric greenhouse gas concentrations, primary vegetative production, and ocean current strength. This record, which stretches back 800,000 years over eight glacial/interglacial cycles in the case of the Antarctica ice core record, shows that the “100,000 year” cycle is not constant but varies between 84,000 and 120,000 years.  The glacial/interglacial cycles are also asymmetric. Warm interglacial periods average only 20,000 years, but cold glacial periods are much longer.

Recent analytic techniques have been able to resolve the paleoclimate record at high resolution. It has been shown that at the beginning of the latest interglacial warming period, about 19,000 to 17,000 years ago, rising temperature in the high latitudes of the Southern Hemisphere preceded rising global CO2 concentration, probably by centuries but by no more than a millennium and a half.  The new analytic tools have revealed that the rise in temperature and atmospheric methane concentration coincided to within decades in Greenland.  The paleoclimate record has revealed that atmospheric methane followed rising CO2 initially, but then plateaued during a northern cooling event (Oldest Dryas), increased sharply during the subsequent northern warming (Bølling-Allerød) period, and dropped precipitously during the final northern cooling period (Younger Dryas) before resuming its rise into the present warmer period (Holocene).  Sea level rose in spurts across the glacial termination, in one case by about 17 meters in less than 350 years.  A new analysis of deep ocean cores has shown that the Northern Hemisphere/Southern Hemisphere “see-saw”, alternating sudden warming in the high north, coinciding with southern cooling and vice versa with a roughly 1,500 year cycle, was associated with and followed variations in the strength of the main Atlantic north/south current, the Atlantic Meridional Overturning Circulation or AMOC.

In this article, the latest contributions to and analyses of the paleoclimate record  elucidate the roles of small changes in Earth’s orbit , the growth of the continental ice sheets, atmospheric greenhouse gases and ocean currents in determining the main features of the observed glacial/interglacial cycles over the past 800,000 years.

Variations in Earth’s orbit and the glacial/deglacial cycle

In 1941 Milutin Milankovitch published a book in which he hypothesized that the cycle of glacial/interglacial periods is controlled by variations in incoming solar radiation (insolation) which are determined by small, predictable changes in the Earth’s orbit and tilt with respect to the sun.  This theory is referred to as orbital forcing and the astronomical cycles that cause this effect are called Milankovitch cycles. According to the orbital theory, changes in summer insolation (heat radiation from the Sun) in the high-latitude Northern Hemisphere were responsible for the glacial/interglacial cycles through their impact on ice-sheet mass balance.

Most of the Earth’s land mass is in the Northern Hemisphere and during ice ages great ice sheets form in high northern latitudes. Summer is when most ice melting occurs. It follows that total summer insolation (heating) at high northern latitudes has been the primary focus of researchers beginning with Milankovitch in trying to determine when interglacials occur. Orbital theory hypothesizes that small variations in the Earth’s orbit affect the amount the heat that the Northern Hemisphere receives during the summer.  Annual insolation can vary by as much as 25%, which the theory hypothesizes affects the summer melting of ice sheets.

Milankovitch cycles showing precession, eccentricity and tilt (obliquity)

The orbital effects that significantly affect insolation are precession, eccentricity and obliquity or tilt. Precession is the most important factor in changing the insolation (solar radiation impinging on the Earth) at high northern latitudes. Precession is the rotation of the Earth’s axis (which is tilted about 23 degrees from the perpendicular) about the perpendicular with a period of about 26,000 years. The Earth’s orbit is not exactly circular (elliptical) which means that the distance between the Earth and the Sun varies. When the Earth is closest to the Sun and the Northern Hemisphere is tilted toward the Sun (summer), northern high latitudes receive more insolation per day, but summers are shorter.

Independently of precession, an increase in obliquity (the angle between Earth’s tilt and a perpendicular to the orbit varies between 22.1 and 24.5 degrees) increases the total amount of insolation received over summer, and this effect is largest at high latitudes. In other words, a greater tilt makes the seasons more extreme in the Northern Hemisphere.

To account for the combined effects of precession and obliquity, Milankovitch used ‘caloric summer half-year insolation’, which represents the amount of energy summed over all summer days which receive more insolation than any day of the winter half.  This is closely approximated by calculating the total summer energy from insolation at 65° N summed over all days for which insolation exceeds 350?watts (W) per square meter. Interestingly at 65° N, precession and obliquity are equally important in determining the variation in the ‘caloric summer half-year insolation’.

Evidence supporting the orbital forcing theory

The first experimental support of the Milankovitch hypothesis was a 1976 article by Hays, Imbrie and Shackleton. It reported the analysis of deep sea sediment cores. It relied on a chronology which was developed by “tuning” sediment timescales to insolation curves calculated from orbital forcing theory. Its conclusions provided some support for the orbital forcing hypotheses, finding the 41,000 and 26,000 year cycles in line with the predicted ratio in their relative strengths in the marine sediment record.  However, it did not find in orbital forcing theory the expected dominant 100,000 year cycle.

The Hays et al. paper has been criticized because it relied on orbital tuning for its chronology.  Orbital tuning is a technique that uses Milankovitch cycles to establish a chronology for dating deep sea sediments.

Comparison of summer insolation at 65 degrees N
Comparison of summer insolation at 65 degrees N (dotted curve) with temperature proxy (solid curve) from paleoclimate record for two glacial terminations. Kyr is thousands of years ago. Source: Raymo et al.

In 1997 a paper by Raymo et al. was the first to provide empirical support for the orbital forcing hypothesis without relying on a chronology determined using orbital tuning.  Delta-oxygen-18 (a temperature proxy) was measured in deep sea sediment cores from eleven sites. The cores came from three oceans, in both high and low latitudes and including eastern and western equatorial regions. Raymo et al developed a common chronology based on recognizable events, typically glacial terminations (deglacial warming leading to deglaciation).  Radiometric measurements based on carbon-14, protactinium-231, and thorium-230 dating were used to determine calendar dates for these recognizable events.

By comparing the delta-oxygen-18 record from the ocean sediment cores with the calculated summer insolation (solar radiation) at 65 degrees N for six of the glacial terminations (denoted TII through TVII), it was found that all the terminations (increasing warming leading to a deglacial period) correlated with periods of increased summer insolation.  It was also noticed that deglacial warming did not always correlate with a period of greatest increased northern insolation indicating that other factors were required to explain the onset of deglaciation.

Evidence against the orbital forcing theory

The first empirical evidence against orbital forcing was in a 1992 paper by Winograd et al. that reported an analysis of calcite cores from Devils Hole (Nevada).  Based on their analysis it was concluded that orbitally controlled variations in solar insolation were not a major factor in triggering deglaciations. The key to the argument was the timing of events at the end of the second glacial period called Termination II (TII). The Devils Hole chronology reported by Winograd et al placed the TII initial warming about 10,000 years before the rise in summer insolation calculated from orbital forcing theory.

In a new 2016 study the original calcite cores from Devils Hole assessed by Winograd et al. have been re-examined and new cores analyzed.  The analysis revealed a systematic offset in the age of calcite deposited at increasing depths in Devils Hole. With this correction the calcite chronology has been reinterpreted as supporting orbital forcing, providing confirmation of the Malinkovitch theory.

The biggest challenge to orbital forcing theory is that it does not by itself predict strong orbital forcing with a “100,000 year” period.  In fact the period is not constant but varies from 84,000 to 120,000 years.  It is also asymmetric – warming is rapid and cooling much slower –  which in not expected from Malinkovitch theory alone.

The “100,000 year” glacial/deglacial cycle

The seminal paper by Raymo et al observed that glacial terminations, which are warming periods leading to deglaciation, did not always correspond to the largest increases in summer insolation calculated from orbital forcing theory. Based on this evidence it was argued that another factor, ice volume, had to be taken into account to explain the “100,000 cycle”.  Once large ice sheets had developed as a result of low temperatures for about 100,000 years, then the first warming of any note caused the ice mass to melt catastrophically, triggering global warming and deglaciation.  It was argued that the interaction between orbital forcing and ice volume were responsible for the pattern of the roughly 100,000 year cycle of glaciation and deglaciation over the last 800,000 years.

Following up on this suggestion, Parrenin and Paillard formulated a mathematical rule based on Raymo’s suggestion that ice volume and increased insolation together trigger deglaciations. Their model hypothesizes that terminations occur when a combination of insolation and ice volume is large. A deglaciation can occur when insolation forcing is moderate if ice volume is very large, or reciprocally when ice volume is moderate if insolation forcing is very large. Parrenin and Paillard’s model is driven by changes in the June Solstice insolation at 65 degrees N and by variations in the Earth’s tilt (obliquity). Parrenin and Paillard report that their simple model not only reproduces sea level transitions at the correct time, but also sea level minima and maxima with the right amplitude.

Recently Tzedakis et al have formulated an empirical rule that requires no input data other than ‘caloric summer half-year insolation’ which can be calculated from astronomical theory. The rule is able to account for the onset of all deglaciations over the past three million years.  It accounts for the dominance of glacial–interglacial cycles with a period of 41,000 years early in the Quaternary and for the change to the “100,000-year cycle” about one million years ago and suggests that the appearance of larger ice sheets over the past million years was a consequence of an increase in the energy threshold required to initiate deglaciation.

Millennial-scale warming and cooling cycles and the role of the Atlantic north/south current

The Northern Hemisphere paleoclimate record reveals that superimposed on the “100,000 year” glacial/interglacial cycle are shorter warming and cooling cycles which extend over 1000 to 2000 years.   The Antarctica ice core record also exhibits millennial-scale warming and cooling events, but they are out of phase with the Northern events and generally lag them. This is an example of the Northern Hemisphere/Southern Hemisphere see-saw. During northern cooling events, ice cores have shown that Antarctica warmed, and each rapid Northern Hemisphere warming was followed shortly by cooling at high southern latitudes.

 

Southern ocean showing surface flow (red), deep water flow (blue), and deep water formation (orange ellipses).

Variations in an important Atlantic current may explain this millennium-scale warming and cooling. The Atlantic Meridional Overturning Circulation (AMOC) brings warm southern surface waters to the North Atlantic where they become saltier and drop to the ocean floor to form North Atlantic Deep Water which flows south ultimately ending up in the Pacific.

Evidence for the role of the Atlantic north/south current has been reported based on an analysis of marine sediment cores.  Between 60,000 and 25,000 years ago in Greenland, during what is called Marine Isotope Stage 3 or MIS 3, there were 15 abrupt warming (8–16 °C) events, each followed by renewed cooling. These cycles lasted on average about a millennium. The most severe of these warming and cooling cycles has been shown to correlate with catastrophic iceberg discharges into the North Atlantic Ocean called Heinrich events. Most of the ice in these events has been shown to be sourced from the Hudson Strait region in modern northern Canada.

One explanation for these abrupt warming events and the hemispheric sea-saw are that they result from changes in the ocean’s persistent circulation pattern.  It is hypothesized that during Northern Hemisphere cold periods the major Atlantic circulation is in a weak or shut-down state. A strengthening or resumption of this current increases heat transport to the North Atlantic and is associated with Northern warming.

A new study in 2016 is the first that provides solid evidence supporting the central role of the Atlantic overturning circulation in triggering Northern warming events. In this study the radioisotope ratio protactinium/thorium (Pa/Th) in microfossil shells (composed of CaCO3) was measured in a core taken from the northwestern Atlantic Ocean.  The Pa/Th ratio is an inverse proxy for the strength of the current at the  core site.

The new results revealed that all Hudson Strait ice discharge events between 60,000 and 25,000 years ago were associated with a dramatic increase in the observed Pa/Th ratio, which is evidence of a major reduction in the Atlantic north/south current.  The results show a persistent pattern of current weakening during cold periods and strengthening during warm periods.   All of the Hudson Strait ice discharge events were found to correlate with the largest reductions in the strength of the Atlantic north/south current.  Further research showed that both sea surface temperature and Greenland temperature proxies consistently lagged the changes in ocean circulation as revealed by the Pa/Th ratio indicating that changes in the Atlantic north/south current preceded Northern Hemisphere climate changes.

Correlation of surface temperature and CO2

Comparison of temperature and atmospheric CO2 from Antarctica Dome C for part 800,000 years. kyr BP is thousands of years before present. Glacial terminations are denoted by TI through TX. MIS is Marine Isotope Stage identifying warming and cooling periods.

The Vostok Antarctica ice core record which extends over the past 420,000 years, the Dome Fuji ice core which reaches back 720,000 years, and the EPICA Dome C ice core which extend back 800,000 years all reveal a close correlation between Antarctica surface temperature and atmospheric CO2 concentration. Newer analytical techniques have been used to analyze the EPICA Dome C ice core at high resolution  over the most recent deglaciation 19,000 to 11,000 years ago.  The high resolution also analysis reveals a close correlation between temperature and CO2 concentration.

The Vostok (Antarctica) ice core record reveals increases in atmospheric CO2 concentration between 80 and 100 parts per million by volume (ppmv) during each of the past four glacial terminations. The ice core record has shown that during the last deglaciation the concentration of CO2 in the atmosphere increased by about 40%.  The resolution of the Antarctica ice core records is not sufficient to determine whether there is a lag between the increases in surface temperature and atmospheric CO2 concentration, but the points at which CO2 and temperature began to rise can be precisely distinguished. The analysis found that the start of rising CO2 lags the rise in Antarctica surface temperature by at least 200 years and at most 1,400 years.

The analysis revealed that the early CO2 increase occurred before any substantial warming in the Northern Hemisphere, which supports the view that the Southern Hemisphere was the source of the CO2 increase.  The results support the hypothesis that the high latitude southern ocean, near Antarctica was the most important factor in regulating the CO2 concentration during the last deglaciation.

Warming and the rise of CO2 during the Last Deglaciation

The last deglaciation has been studied intensively to try to understand how interglacial global warming was triggered and how it developed. The global climate underwent massive changes from the end of the Last Glacial Maximum (LGM) to the modern warm period (Holocene) from 19,000 to 11,000 years ago. During the LGM temperatures in East Antarctica were approximately 9–10 °C lower than today. In Greenland average temperatures were 15 °C lower.  Atmospheric CO2 dropped to 180–190 ppm.  Since atmospheric CO2 is the carbon source for C3 plant photosynthesis, this likely had a major impact on global plant productivity.  At the end of all glacial periods atmospheric dust increased.   Atmospheric CO2 concentrations increased by 80 to 100 parts per million by volume (ppmv) across glacial terminations.   At the end of the last ice age, about 17,500 years ago, CO2 and methane concentrations began to increase concomitantly and reached interglacial maximum levels around 11,700 years ago.

 

 

In a recent study sea surface temperatures and precipitation from ice cores, sea floor sediments, pollen, cave calcite records, and sea phytoplankton records were compiled and analyzed using a temporal-statistical technique called principal component analysis. Two important trends were identified.  The key trend is a global warming trend that started at a time of increasing Northern Hemisphere summer insolation and ultimately brought temperatures to pre-industrial levels. This trend correlates strongly with increasing greenhouse gas levels. The second trend responsible for the millennial-scale warming and cooling periods such as the Oldest Dryas, Younger Dryas and intervening Bølling-Allerød period.

A recent study has used a new analytical technique to measure temperature and other key variables in Greenland ice cores with less than yearly resolution for the first time. The study analyzed the two warming and one cooling events in the period from 15,500 to 11,000 years ago. The high-resolution records reveal that the two warming events involved a warming from glacial to warm interglacial of more than 10 °C. The warming transition beginning 14,700 years ago at the end of the Oldest Dryas and leading into the Bølling-Allerød warming period occurred within only three years.  Associated with this was with a sea level rise of 12-22 meters within a period of a maximum of 350 years.  The other warming transition 11,700 years ago at the end of the Younger Dryas cooling period leading into the Holocene was also rapid, occurring over 60 years.  The high-resolution records suggest that the two warming events followed the same general pattern and involved a sudden shift over one to three years in polar atmospheric circulation.

The second-to-last warming event of the last glaciation, at the end of the Oldest Dryas, initiated the Bølling–Allerød warming. There is clear evidence for this transition in numerous palaeoclimate records, but the relative timing of climate shifts in different regions of the world is not clear.  Recently, the phasing of global climate change at the onset of the Bølling–Allerød event has been investigated using the North Greenland Eemian (NEEM) ice core. Specifically, methane concentrations, which act as a proxy for low-latitude climate, and the nitrogen-15/nitrogen-14 ratio of atmospheric nitrogen, which reflects Greenland surface temperature, were measured over the same interval of time.

 

 

The results show that methane emissions and Greenland temperature changed essentially synchronously.  Since there are no known methane sources in the immediate vicinity of Greenland, the synchronous changes in temperature and methane emissions revealed rapid transmission of the abrupt Bølling–Allerød climate change over a large geographical area.  At this time scale (decades) the interpretation of high-resolution ice core data is complicated by uncertainties in estimating methane mixing in the atmosphere as well as diffusion of gases in the snow and ice (firn) of the Greenland ice sheet. These complications imply that ice core gas records are difficult to interpret at time intervals of less than decades.   An uncertainty analysis found that methane sources could have led Greenland temperature by up to 24 years.  Alternatively, temperature could have led methane by up to 21 years.

On the other hand, there is strong evidence that southern ocean warming pre-dated the rise of atmospheric CO2. Radiocarbon dating of micro organisms living on the deep ocean floor and in surface waters in a marine core collected in the western tropical Pacific has been used to determine the relative chronology of warming in the southern ocean near Antarctica and rising CO2 during the last deglaciation. The results provide evidence that at high and medium latitudes the southern ocean warmed by about 2°C between 19,000 and 17,000 years ago, about 1,000 years before the rise in atmospheric CO2.

The early warming in the Antarctic indicates that the mechanism responsible for initiating the deglaciation did not begin in the tropics nor was it initiated by CO2 forcing. Both CO2 concentration and the tropical sea surface temperatures did not begin to change until after 18,000 ago, approximately 1,000 years after the deep water record indicates that the southern ocean was warming.  It has also been suggested that a possible trigger for the initial deglacial warming around Antarctica was the increase in solar insolation over the southern ocean during the Southern Hemisphere spring.

 

 

The period of the first pulse of CO2 into the atmosphere lasted from about 17,000 to 14,500 years ago and accounted for about 35% of the total increase in CO2 during the deglaciation.  Research has shown that during this interval there was a dramatic drop in delta-carbon-13.  There was also a significant drop in radiocarbon beginning at 17,000 years ago and ending with the Bølling/Allerød warming period at about 14,500 years ago.   One author referred to the coeval drop in delta-carbon-13 and delta-carbon-14 during the period 17,500 to 14,000 years before the present asarguably the most enigmatic carbon cycle change in the course of the [deglacial] transition.”  For this reason the period has been dubbed the “Mystery Interval”.

However, an account of what occurred during this period is beginning to emerge. The drop in atmospheric delta-carbon-13 suggests a large infusion of CO2 into the atmosphere from an organic source.  The drop in delta-carbon-14 is indicative of an organic carbon pool that had been isolated from the atmosphere for a long period of time, certainly from before 35,000 years ago.  The accelerated deposition of biogenic opal observed during this period indicates a period of upwelling in the southern ocean.  Research has also found evidence of carbon-14 depletion in the southern ocean deep water during the last ice age. This body of carbon-14 depleted water gradually disappeared during the “Mystery Interval” as mixing broke down ocean stratification.  In the north the rise in atmospheric methane suggests expanding and more productive wetlands in northern and equatorial regions.  It has also been suggested that melting permafrost may have also contributed to the rise in atmosphere CO2.

Global surface temperature and CO2

Location of sites used to calculate paleoclimate global surface temperature
Location of sites used to calculate paleoclimate global surface temperature. Histogram shows coverage by latitude. Colours indicate different types of temperature proxy.

Global surface temperature is a derived quantity, computed by averaging temperatures recorded at different locations on the Earth’s surface.  It can be affected by a number of factors relating to sampling locations and the temperature proxies used.  Nevertheless, to better understand the global role of CO2 during the last deglaciation, an attempt was made to reconstruct global surface temperature using paleoclimate temperature proxy records from 80 geographically distributed sites.

Comparing these surface temperatures with the atmospheric CO2 record over the same time period reveals that rising atmospheric CO2 is correlated with, but leads global surface temperature during the last deglaciation.

 

 

The database of temperature proxies from 80 sites shows that changes in temperature at all latitudes, as well as a net global surface temperature increase of about 0.3 °C, preceded the initial increase in CO2 concentration at 17,500 years ago. This strongly suggests that rising CO2 did not initiate deglacial warming.

Comparison of global surface temperature, Antarctica temperature, and atmospheric CO2
Comparison of global surface temperature, Antarctica temperature, and atmospheric CO2. Global surface temperature (blue), Antarctic ice-core temperature record (red), and atmospheric CO2 concentration (yellow dots). The Holocene, Younger Dryas (YD), Bølling–Allerød (B–A), Oldest Dryas (OD) and Last Glacial Maximum (LGM) intervals are indicated. Source: Shakun et al.

The database reveals that global warming before 17,500 years ago occurred in two phases. There was a gradual increase between 21,500 and 19,000 years ago, and then a steeper increase between 19,000 and 17,500 years ago. The data shows that the first increase is associated with warming of the northern mid to high latitudes, particularly in Greenland.

 

 

The second increase occurred during a Northern/Southern hemisphere seesaw event, possibly associated with a reduction in Atlantic north/south current strength as seen in the Pa/Th record. Tropical and Southern Hemisphere warming occurred coincidentally with northern cooling. The warming southern ocean led to rising atmospheric CO2. The rising temperature led rising CO2 by at most 800 years. The rising CO2 levels drove global warming as seen in comparing it to the reconstruction of global surface temperature.

One of the findings revealed by the analysis of Antarctic ice cores is that every glacial maximum has been characterized by the same low atmospheric CO2 concentration and low surface temperature. The repeated occurrence of the same low CO2 concentration and surface temperature prior to deglaciations is suggestive of a feedback mechanism that kicks-in whenever climactic conditions approach a threshold. A recent study argues that a biological feedback mechanism may be responsible for the rapid increase in atmospheric CO2 and temperature leading to deglaciation at the end of ice ages.

Plausible account of the last deglaciation

Orbital forcing as the trigger for the “100,000 year” glacial/interglacial cycle is now widely accepted within the climate science community. However, as Raymo pointed out, orbital forcing by itself does not explain the major events of the Earth’s climate over the past 800,000 years.  Hence, the most plausible explanation for the observed “100,000 year” cycle takes into account ice volume in addition to orbital forcing.  It has been proposed, and a mathematical rule has been formulated which supports the empirical evidence, that once a large ice sheet has developed (which requires low temperatures in the Northern Hemisphere for about 100,000 years), the first warming of any note attributed to orbital forcing, causes the ice mass to melt catastrophically, triggering global warming and the interglacial cycle.

A plausible depiction of the sequence of events has emerged that explains most of the variation in the Earth’s temperature and CO2 during the last deglaciation.  At the end of the last ice age 21,500 and 19,000 years ago,  Northern Hemisphere warming  caused by increased summer insolation began to melt the massive ice sheets which covered much of the Northern Hemisphere.  Between 19,000 and 17,500 years ago the Atlantic north/south (AMOC) current was disrupted and the hemispheric sea-saw mechanism resulted in Northern Hemisphere cooling and Southern Hemisphere warming in the southern ocean, Antarctica, and high elevation glaciers.  Between 19,000 and 17,000 years ago the southern ocean off Antarctica warmed by ~2°C.  This Southern Hemisphere warming may have been augmented by rising spring insolation in high southern latitudes arising from orbital forcing.

Southern ocean showing current directions and location of deep water formation.
Southern ocean showing current directions and location of deep water formation. Blue boxes denote ice and sea sediment coring locations.

About 800 years after the southern ocean warming atmospheric CO2 began to rise.  The period of the first pulse of CO2 into the atmosphere lasted from about 17,000 to 14,500 years ago and accounted for about a third of the total increase in CO2 during the deglaciation.  During this period there is evidence that increased ocean mixing and the break down of southern ocean stratification released a large amount of old (carbon-14 depleted) carbon from the southern ocean deep water. In the north the rise in atmospheric methane suggests expanding and more productive wetlands in northern and equatorial regions.  It has also been suggested that melting permafrost may have also contributed to the rise in atmosphere CO2.

The CO2 ventilated from the southern ocean was probably the major driver during the early period of the deglaciation, perhaps in combination with CO2 generated in the north.  There is uncertainty about the source of the CO2 that was responsible for the remaining 65% increase in atmospheric CO2 during the rest of the deglaciation. Whatever its source, it is generally believed that rising CO2 amplified the warming trend driving the increase in the Earth’s global surface temperature.  Sea level rise rose by 80 meters, but in spurts, in one case of 17 meters in less than 350 years.  In the north rising temperature increased the extent and productivity of northern and tropical wetlands thus increasing atmospheric methane.  There is also evidence that the climates in the North Pacific and the North Atlantic, which were typically out of phase, were in phase throughout the deglaciation.

Recent research is also able to show that the millenium-scale warming and cooling events in the north, which are observed superimposed on the broader “100,000 year” cycle, are associated with variations in the Atlantic north/south current  or AMOC.  These include the warming events at the ends of the Oldest and Younger Dryas, which occurred rapidly, over decades and even years, and the rapid cooling event at the end of the Bølling–Allerød.  At the same time in the Antarctic, when the north cooled the southern hemisphere experienced warming and vice versa, the result of what has been dubbed the hemispheric see-saw.  The warming period at the beginning of the Bølling–Allerød period is associated with an abrupt rise in sea level of between 12 and 22 meters over 350 years or less.

Further research

 

 

Comparison of temperature proxies and CO2 concentration
Evidence for repeated coincident CO2 and surface temperature minima from the ice core record (a) Temperature proxy – Delta-oxygen-18 from surface plankton (b) Temperature proxy – Delta-D from EPICA Dome C Antarctica, (c) CO2 concentration from multiple ice core analyses. Horizontal dashed lines indicate the average minima of delta-D and CO2.

This account of the last deglaciation raises questions.  The key question is what were the physical and biological mechanisms which caused this sequence of events.   A related question is what was the  mechanism that maintained temperature and atmospheric greenhouse gases in lock step over the past millions of years.

An intriguing observation is that the Antarctica ice core record reveals that at each glacial maxima, surface temperature and CO2 concentration dropped to the same minima.   The repeated occurrence of the same low CO2 concentration and surface temperature prior to deglaciations is suggestive of a feedback mechanism that kicks-in whenever climactic conditions approach a threshold.  Several mechanisms have been proposed to account for this.

The ice core record shows that the CO2 concentration at these minima is about 190 ppm. There is independent experimental evidence that CO2 concentrations at this level begin to inhibit photosynthesis.  When applied to the entire Earth ecosystem, the physiological responses imply large reductions in Net Primary Productivity or NPP (the net carbon uptake by plants after accounting for plant respiration) and carbon storage during glacial periods.  Laboratory experiments have found that marine nitrogen fixing bacteria (cyanobacteria) are especially sensitive to low CO2, with growth rates strongly decreased when CO2 concentration drops below 200 ppm.  If nitrogen fixation was progressively handicapped as CO2 fell because nitrogen fixers were having a hard time growing, it could have inhibited the soft tissue pump, shifting carbon from the ocean back to the atmosphere. In support of this mechanism, at the end of the last ice age, there is evidence of CO2 degassing from the Southern Ocean.

A mechanism that has recently been suggested is a feedback system involving low CO2 and temperature resulting in increased atmospheric dust and greater absorption of solar radiation. During the initial glacial period, the high reflectivity of the northern ice sheets reflects most of the solar radiation resulting in cooling. As the oceans and atmosphere cool, more atmospheric CO2 is absorbed by the oceans. Atmospheric CO2 concentrations eventually reach a critical minimum of about 190 ppm, which combined with cool arid conditions, cause a die-back of temperate and boreal forests and grasslands, especially at high latitudes. The ensuing soil erosion generates dust storms, resulting in increased dust deposition on the northern ice sheets and greater absorption of solar radiation. As northern hemisphere solar radiation increases during the next Milankovitch cycle, the dust-laden ice-sheets absorb more solar radiation and undergo rapid melting, which forces the climate into an interglacial period. In support of this mechanism, Antarctic ice cores provide evidence of increasing atmospheric dust at the end of all ice ages over the past 800,000 years.

Further reading

(1) Empirical evidence for orbital forcing theory

For more than a century the cause of fluctuations in the Earth’s climate responsible for the growth and ebb of the great ice sheets has remained an unsolved scientific mystery. Only the hypothesis that relates these changes to small variations in the Earth’s tilt and orbit changing the amount of solar radiation hitting the Earth has been formulated so as to predict the frequencies of major glacial/deglacial cycles. This seminal 1976 paper tested this hypothesis by treating changes in the Earth’s orbit and tilt as a forcing function of a system whose output is the 450,000 year geological record of the climate as recorded in marine sediments. The geological data comprise measurements of three variables related to climate in two deep-sea sediment cores. The analysis provides some support for the orbital forcing hypothesis, but reveals orbital forcing theory by itself is not able to explain the dominant 100,000 year glacial/deglacial cycle.

What causes the ebbing and waxing of Quaternary ice sheets ?

(2) Further evidence supporting orbital forcing theory, but suggesting that accounting for the 100,000 year glacial/deglacial cycle requires considering the extent of the continental ice sheets in addition to orbital forcing

The first experimental support of the orbital forcing hypothesis was published by Hays, Imbrie and Shackleton in 1976.  The chronology of the sea sediment cores was established using orbital tuning.  Using this chronology to support the orbital forcing hypothesis has been criticized because it was developed by tuning sediment timescales to insolation curves calculated from the astronomical effects studied by Milankovitch. In 1997 Raymo was the first to provide observations supporting the orbital forcing hypothesis without relying on orbital tuning.  Raymo also noticed that ice age terminations occurred only after considerable build-up of ice sheets, and that beyond this point, the next northern latitude summer insolation maximum, even a relatively weak one, appeared to trigger deglaciation.

Are Milankovitch cycles responsible for the glacial/interglacial periods of the past 800 000 years ?

(3) Mathematical rule for estimating the period of glacial/deglacial cycles and the associated sea level rise based on orbital forcing and ice volume of the continental ice sheets.

The intervals between glacial terminations over the past million years range form 84 kyr to 120 kyr, or roughly every 100 kyr. But the variation in insolation computed from the astronomical model does not does not predict a strong signal at this frequency. Raymo noticed that terminations occurred only after considerable build-up of ice sheet, and that beyond this point, the next northern latitude summer insolation maximum, even a relatively weak one, will cause deglaciation. To support this hypothesis, Parrenin and Paillard suggest that ice volume and insolation together play a role in triggering deglaciations and that terminations occur when a combination of insolation and ice volume is large. More precisely, a deglaciation can occur when insolation forcing is moderate if ice volume is very large, or reciprocally when ice volume is moderate if insolation forcing is very large.

Glacial/interglacial cycles determined by Milankovitch insolation cycles and ice volume

(4) Empirical evidence against orbital forcing theory

In this historically important paper, delta-oxygen-18 variations in a 36-centimeter-long core of vein calcite from Devils Hole, Nevada, provide a continuous 500,000-year paleotemperature record that exhibits all the major features in marine ice-volume and the Vostok (Antarctica) paleotemperature records. Comparing the timing of glacial terminations with calculated Northern Hemisphere insolation revealed that orbitally controlled (Milankovitch cycles) variations in solar insolation were not a major factor in triggering deglaciations.

Devils Hole record 4,500 to 560,000 years ago does not support orbital forcing hypothesis

(5) Revised analysis of Devils Hole calcite cores provides support for orbital forcing theory

In 1976 an analysis of deep sea sediment cores provided the first evidence supporting this theory. However, an influential study of calcite cores from Devils Hole in 1992 concluded that the evidence did not support the theory. In this new 2016 study the original calcite cores from Devils Hole were reexamined and new cores analyzed. A systematic correction was identified that is required in dating calcite cores in Devils Hole and with this correction the calcite chronology does support the theory, thus vindicating the orbital forcing hypothesis.

New evidence from Devils Hole affirms interglacial warming linked to small variations in Earth’s orbit

(6) Vostok (Antractica) ice core record reveals close correlation between temperature and CO2

The Vostok ice core was the first to extend the paleoclimate record including temperature and atmospheric gas composition back several hundred thousand years.

Vostok correlation of temperature and greenhouse gas concentrations during the last 420,000 years and four glacial-interglacial cycles

(7) Dome Fuji (Antractica) ice core shows strong correlation between atmospheric CO2 and temperature

The Dome Fuji ice core stretches back about 720,000  years.  The delta-oxygen-18  and other radioisotope records are part of the ice itself.  The ice incorporates air bubbles which can be analyzed to determine atmospheric composition of CO2 and other gases.  However, because of the physical process by which snow and ice gradually trap air bubbles from the atmosphere, the air bubbles and the enclosing represent different ages.  A physical model of this process is required to accurately date the air bubbles relative to the enclosing ice.

Dome Fuji ice core sampling shows strong correlation between CO2 concentration and temperature

(6) EPICA Dome C (Antractica) ice core climate record reveals close correlation of atmospheric CO2 and temperature over past 800,000 years.

The EPICA Dome C ice core extends the paleoclimate record including temperature and atmospheric gas composition further back than any other ice core.

Dome C ice core climate record for the past 740,000 years and eight glacial-interglacial cycles

(9) High resolution analysis of Dome C ice core shows temperature rise in Southern Hemisphere high latitudes preceded CO2 and CH4 rise

This important article provides evidence that southern ocean warming pre-dated the rise of atmospheric CO2. Radiocarbon dating of micro organisms living on the deep ocean floor and in surface waters in a marine core collected in the western tropical Pacific has been used to determine the relative chronology of warming in the southern ocean near Antarctica and rising CO2 during the last deglaciation. The results provide evidence that that the southern ocean off Antarctica warmed by ~2°C between 19 and 17 thousand years before the present (ky B.P.), about one thousand years before the rise in atmospheric CO2.

Southern Ocean warming preceded rising CO2 during the last deglaciation

(10) First concrete evidence that the changes in the Atlantic north/south current triggered the millennium-scale Northern Hemisphere warming and cooling events.

It has been suggested that variations in the Atlantic Meridional Overturning Current (AMOC), which draws southern tropical surface waters to the North Atlantic where they become increasingly saline and sink to become deep water and flow south, has played a central role in abrupt Greenland warming (8–16 °C) events and cooling periods during the last ice age. A new study reports measurements from ocean floor sediments that provide the first direct evidence that changes in ocean currents have triggered abrupt northern hemisphere warming.

New evidence that changes in ocean currents triggered abrupt Northern hemisphere warming

(11) Orbital forcing, atmospheric CO2 and ocean currents account for temperature variability during last deglaciation

In this study the authors compiled and analyzed sea surface temperatures and precipitation from ice cores, sea floor sediments, pollen, cave calcite records, and sea phytoplankton records.  A  principal component or empirical orthogonal function (EOF) analysis was used to decompose geopatial-temporal data sets to separate physical processes from background noise. The most important trend that the analysis revealed is a global warming trend that started at about the time of increasing Northern Hemisphere summer insolation and ultimately brought temperatures to pre-industrial levels. This trend correlates strongly with increasing greenhouse gas levels. The other important trend the analysis revealed is responsible for the pronounced millennial-scale variability such as the Oldest Dryas, Younger Dryas and intervening Bølling-Allerød period. This trend is associated with changes in the Atlantic meridional overturning circulation (AMOC).

Accounting for global and regional warming during the last deglaciation

(12) High resolution analysis of Greenland ice cores reveals abrupt temperature changes of 10 °C over three years 2008

This high-resolution study provides important new insights into both the onset and evolution of abrupt climate change. Using a new analytical technique, temperature and other key variables have been measured in Greenland ice cores at better than yearly resolution for the first time. The study analyzed three abrupt climate events (two warming and one cooling) in the period from 15.5 to 11 thousand years ago. The high-resolution records reveal that the two warming events involved a warming from glacial to warm interglacial of more than 10 °C. The warming transition beginning 14,700 years ago occurred within only three years. The other warming transition 11,700 years ago extended over 60 years. The high-resolution records demonstrate that the two warming events followed the same general pattern involving a sudden shift over one to three years in polar atmospheric circulation.

High resolution analysis provides important new insights into abrupt climate change

(13) High resolution analysis of a rapid warming event during the last deglaciation 2014

It has been difficult to determine whether warming is led by temperature or by rising methane concentration. In this article the authors use measurements from Greenland ice cores for a single warming event to conclude that warming in the tropics and Northern latitudes occurs nearly synchronously. They also find increasing temperature and methane emissions begin at the same time. Uncertainties in dating bubbles in ice cores prevent determining with confidence whether temperature or increasing methane emissions led the warming. Within the estimated error bars, temperature could have led increasing methane emissions by up to 21 years, or alternatively methane emissions could have led increasing temperature by up to 24 years.

How does rapid climate change occur: The Greenland ice core record

(14) Melting permafrost and rising atmospheric CO2

During the beginning of the last deglaciation between 17,500 and 14,500 years ago atmospheric CO2 concentrations began rising from about 190 ppm in glacial times to approximately 270 ppm by the beginning of the present warm period. There is evidence that the rise in CO2 was from an old carbon pool that was formed from a biological source.   Two possible sources are the southern ocean and permafrost soils.  A large amount of carbon, of the same order of magnitude as is contained in the atmosphere, was deposited in permafrost soils during the last ice age.  In addition the oceans contain about ten times as much CO2 as the atmosphere.  It has been suggested that if melting permafrost did initially contribute to rising CO2 levels during the last deglaciation, the ventilation of CO2 from the oceans must have amplified this effect.

Estimating greenhouse gas emissions from melting permafrost

(15) Rising CO2 correlates closely with and leads increasing temperature during the last deglaciation

Global surface temperature during the last deglaciation is reconstructed from 80 geographically distributed proxy records. Comparing this with the CO2 record reveals that rising atmospheric CO2 is correlated with and leads global surface temperature during the last deglaciation. Superimposed on this dominant trend the authors find that a millenial-scale seasawing of temperature changes of the Northern Hemisphere and Southern Hemisphere correlates with the strength of the Atlantic meridional overturning circulation (AMOC) recorded in marine sediments.

CO2 drove, but did not trigger, warming during last deglaciation

(16) Greenhouse gases amplify the effect of orbital forcing on the Earth’s climate

This article assumes that the ultimate driver of glacial-interglacial transitions has been the weak influence of varying insolation (solar radiation) associated with Milankovitch cycles (changes in the Earth’s orbital geometry). The Vostok ice cores from Antactica have shown a close, positive correlation between temperature and CO2 and CH4 concentrations in the Earth’s atmosphere over the past 400,000 years. This article makes the argument that CO2 and CH4 concentrations amplify the weak influence of changes in the Earth’s orbit and are responsible for much of the observed paleoclimatic variation during the past million years.

Greenhouse gases amplify the effects of changes in the Earth’s climate

(17) Coral cores reveal an increase in sea level of 14-18 meters occurred in less than 350 years

During the last deglaciation there were several episodes of rapid climate change including dramatic periods of sea-level rise.  One warming period, referred to as  MWP-1A, led to sea level rise of about 20?meters,in less than 500?years. This has been associated with a massive collapse of ice sheets and a massive meltwater pulse into the North Atlantic. Based on corals drilled offshore from Tahiti, the authors found that MWP-1A started no earlier than 14,650 years ago and ended before 14,310 years ago. Their observations from analysis of corals show that the increase in sea level at Tahiti was between 12 and 22 metres, with a most probable value between 14 and 18 meters over a period that does not exceed 350 years, but could be less than a century. This implies that the rate of global sea-level rise exceeded 40 millimeters per year during MWP-1A.

Evidence for rapid sea level rise during the last deglaciation

(18) High resolution analysis of Antarctica ice cores reveals close correlation of temperature and atmospheric CO2 with temperature beginning to rise 800 ± 600 years before CO2

The Vostok (Antarctica) ice core record, which extends over past 420,000 years, shows increases of the CO2 concentration between 80 and 100 ppmv for each of the past four glacial terminations. During the last deglaciation the concentration of CO2 in the atmosphere increased by ?40%. In this study the EPICA Dome C (Antarctica) ice core has been used to measure a high resolution record of CO2 and CH4 concentration over the last deglaciation. Comparing the CO2 record to the deuterium abundance or ?D of the ice (proxy for surface air temperature) reveals a close correlation. The resolution of the record is not sufficient to determine whether there is a lag between the two variables, temperature and CO2 concentration. The points at which CO2 and temperature began to rise were estimated to be 17,000 ± 200 years for the start of the CO2 increase and 17,800 ± 300 years for the start of the ?D increase. The start of the CO2 increase was estimated to lag the temperature rise by 800 ± 600 years. This is consistent with a dominant role of the Southern Hemisphere in causing the CO2 increase. The methane increase, on the other hand, is thought to have been caused by an expansion of wetlands in the tropics and northern latitudes.

CO2 lagged rising temperature in the Southern Hemisphere during last deglaciation

(19)  North Atlantic and North Pacific climates, which often behave as a see-saw,  were synchronized during last deglaciation 15,500 to 11,000 years ago.

By comparing the temperature record in the North Pacific and the North Atlantic, the authors of this study find that over most of the past 18,000 years, the relationship between sea temperatures in the North Atlantic and North Pacific was often a seesaw. Warm temperatures in the Pacific often correspond to cold temperatures in the North Atlantic and vice versa. The authors find that the occasional dynamic coupling of North Pacific and North Atlantic climates may be linked to critical, but poorly understood, transitions in Earth’s climate system such as the onset of deglacials. As evidence the authors find that about 15,500 to 11,000 years ago the climate in the two regions synchronized, meaning similar temperatures in both basins, just prior to and during the most abrupt climate transitions of the last 20,000 years and just prior to the current deglacial warm period. This conclusion supports the notion that to understand Earth’s warming and cooling cycles requires a combination of external forcing such as solar radiation and the internal dynamics of Earth’s climate system including its atmosphere, oceans and ice sheets.

Synchronization between the North Pacific and North Atlantic and the onset of deglacial warming

(20) Compilation of the latest  paleoclimate data show that changes in atmospheric temperatures and greenhouse-gas concentrations can be determined with very high confidence from polar ice cores.  Since AR4 these records have been extended from 650,000 to 800,000 years ago.

Past changes in atmospheric temperatures and greenhouse-gas concentrations can be determined with very high confidence from polar ice cores. Paleoclimate data provide quantitative information about the Earth’s climate system response to external factors such as changes in solar radiation, volcanic activity and Earth’s orbit with respect to the Sun. It helps understand Earth’s climate system on timescales longer than a few centuries and transitions between different climate states which occurred on timescales of decades to a few centuries to milennia. Past changes in atmospheric temperatures and greenhouse-gas concentrations can be determined with very high confidence from polar ice cores. Since AR4 these records have been extended from 650,000 years to 800,000 years ago.

IPCC AR5: Earth surface temperatures, greenhouse gas concentrations, sea level rise and ice loss from the paleoclimate record

(21) Evidence that Southern Ocean ventilation contributed to CO2 rise during last deglaciation

The radiocarbon record reveals that radiocarbon-depletion in deep water, which originated in the last ice age, disappeared during the first 2000 years of the last deglaciation. This is consistent with a Southern Ocean outgassing event that resulted in a significant drop in atmospheric radiocarbon between 17,000 and 14,500 years ago.

Evidence for CO2 outgassing from the Southern Ocean at the beginning of the last deglaciation

(22) Fingerprinting the source of CO2 during the last deglaciation

During the last deglaciation, about 20,000 to 12,000 years ago, the concentration of CO2 in the atmosphere increased from about 190 ppm to 270 ppm. In this study the stable carbon isotope ratio (ratio of carbon-13 to carbon-12) in the atmosphere is used to fingerprint the source of the CO2 increase. Since plants preferentially absorb the lighter isotope carbon-12, the stable isotope ratio is less in plants than in the atmosphere. In this study the stable carbon isotope ratio is reconstructed over the past 24,000 years from Antarctic ice cores. The time series reveals that during the first period of increasing atmospheric CO2 from 17,000 to 15,000 years ago the stable isotope ratio dropped precipitously indicating that a source of the CO2 was a large pool of carbon-13 depleted carbon of organic origin. There is independent evidence that this old carbon resulted from upwelling of deep water in the Southern Ocean.

Fingerprinting the source of rising CO2 during the last deglaciation

(23) Evidence of biological feedback driving warming

One of the findings revealed by the analysis of Antarctic ice cores is that every glacial maximum has been characterized by the same low atmospheric CO2 concentration and low surface temperature. The repeated occurrence of the low CO2 and surface temperature may be the result of a feedback mechanism that kicks-in whenever climactic conditions approach a threshold. Extensive research has shown that photosynthesis is inhibited by low atmospheric CO2. A recent study argues that a biological feedback mechanism may be responsible for the rapid increase in atmospheric CO2 and temperature leading to deglaciation at the end of ice ages.

Evidence of biological feedback driving warming at ice age terminations

(24) Simple rule accounts for onset of ice age deglaciations over the past three million years

A recent study has created a simple mathematical rule that can account for the timing of the onset of interglacials following ice ages over the past three million years. The rule is based on predictable long-term astronomical variations in the Earth’s orbit and tilt called Milankovitch cycles, without any knowledge of atmospheric greenhouse gas concentrations, ice sheet dynamics, volcanism, cosmic rays, dust or other climate data.

Simple rule accounts for onset of ice age deglaciations over the past three million years

(25) Cooling preceded shift to 100,000-year glacial/deglacial cycles 800,000 years ago

In this study the global average surface temperature over the past 2 million years has been derived from deep sea cores using a newly developed methodology. It was found that the Earth’s surface temperature gradually cooled until 1.2 million years ago after which it has remained stable when averaged over glacial/interglacial cycles. The results reveal that global cooling occurred about 300,000 years before the rapid ice sheet growth and the development of the first 100,000-year glacial/deglacial cycle about 800,000 years ago. This suggests that global cooling was a key factor, but not the sole cause, in the shift to 100,000-year glacial cycles.

Global surface temperature reconstruction reveals cooling preceded shift to 100,000-year glacial/deglacial cycles 800,000 years ago