Global Warming in the Context of Glacial Cycles


Using the Column Radiation Model, GFS analysis data for 2010, CERES derived surface albedo, and Last Glaical Maximum (LGM) ice and land mask estimates, it is possible to compare resulting net radiance for a number of scenarios ( with a large number of assumptions and unknowns). From these comparisons:

  • CO2 Doubling indicates a large change in global mean radiative forcing(RF)
  • The glacials and inter-glacials indicate relatively little change in global mean RF
  • The glacials and inter-glacials appear to be a result of max RF, not global mean
  • CO2 Doubling RF change is greatest at the equator, least at the poles
  • The LGM RF gradients had the highest mean but lowest range
  • The interglacial RF gradients exhibited the greatest range
  • Both factors are consistent with dry glacials and wet interglacials
  • The RF gradients for a CO2 Double do not indicate much change
  • The RF gradients for a CO2 Double with no sea ice do indicate a summer reduction


In the course of discussing global warming, many questions and comparisons arise with the glacial and interglacial periods. I wondered how global warming might compare and contrast with these periods. Using the column radiation model for a given GFS data set ( the 21st day of each month for the year 2010, at 0Z,6Z,12Z,and 18Z ) it is possible to compare some idealized scenarios. What follows are the results of these runs. There are many assumptions, variances, and unknowns that create uncertainty with these results. However, the relative importance of incoming radiation from orbits, and the relative portion of greenhouse gas forcing from the global warming case identifies some

Caveats, Assumptions, Exclusions

The cases considered are listed in the table below.

All cases use the 2010 atmosphere, including surface elevations, temperature, humidity, and clouds. Temperature is both the result of and cause of radiance change ( in the infrared ). Presumably, since constant 2010 temperatures are used for all scenarios, the effect is to exaggerate somewhat the longwave influences. Also, in these runs, a constant aerosol value is used. The effect of sea level rise is imposed in changing the land/sea mask, but not on  atmospheric levels. Similarly, the last glacial maximum incurred a 120m sea level fall, not accounted for here, which changed the distribution of the atmosphere, and made the atmosphere optically thicker over the oceans and optically thinner over the continents. Also, ocean albedos and sea ice albedos are the same over the scenarios, varying only as a result of solar elevation. Also, the land albedos ( where applicably masked ) are from the modern CERES estimate. Unglaciated land albedo is thought to have been higher during the glacials and lower during the inter-glacials, but the CERES albedos are used throughout.

Table of scenarios:

Name Year(CE) CO2    ppm CH4  ppm N2O   ppm CFC11   ppt Land Albedo Sea Ice
EEMIAN -126000 280 600 200 0 CERES  &4mSLR  2010   -10%
LGM -24500 190 400 200 0 CERES &NSIDC NSIDC
HCO -5000 270 600 200 0 CERES  &4mSLR 2010   -10%
PI 1750 279 700 270 0 CERES 2010
Recent 2010 400 1800 324 240 CERES 2010
Double 2100 800 1800 324 240 CERES  &4mSLR 2010  -10%
HalfSI 2100 800 1800 324 240 CERES  &4mSLR 2010  -50%
NoSI 2100 800 1800 324 240 CERES  &4mSLR NONE
Next 11000 279 700 270 0 CERES  &4mSLR 2010  -10%

Comparison with CERES

An animation of the runs ( average of the 0Z,6Z,12Z and 18Z observations for the 21st of each day of the month ) yields the average net radiance versus CERES monthly average data below.


The CRM calculations on the twelve days in 2010 provide qualitative agreement with the CERES monthly averages, but there are significant variances accounted for both by single days in 2010 versus monthly averages for fifteen years, but also due to the assumptions above and perhaps also due to the CRM model which is now quite old. Consequently, the results below are likely not realistic absolute estimates. However, within the many assumptions and limitations, the runs provide useful relative comparisons.

 Averages versus Extremes

The focal point about global warming is the global mean surface air temperature. The radiative forcing imposed by an increase in greenhouse gases is quite apparent observing the difference between annual average Double scenario with that of PreIndustrial:


The Double scenario ( green ) exhibits the greatest annual average difference from PreIndustrial for all scenarios and at all latitudes. Other features of note:

  • radiative forcing(RF) due to CO2 doubling is greatest at the equator, least at the poles
  • radiance at the LGM was the lowest in the annual average
  • radiance from LGM was lowest at the glacial latitudes ( both Northern and Southern )
  • radiance from the Eemian and HCO did not differ much from PreIndustrial
  • radiance from the Eemian and HCO were lower than PreIndustrial in the tropics

The glacial region minima for the LGM scenario is consistent with the ice albedo feedback.

The difference of global average radiance from the PreIndustrial scenario results in the following comparison:


This comparison is often argued to reflect the relative importance of RF to climate. The large deficit during the LGM, the relative increase ( correlating with warming ) through 2010, and the the future increase with a Double of CO2 are generally consistent with this idea. However, when considering the HCO, the Eemian, or the Next Arctic orbital maxima, the relationship of global annual average radiance and climate regimes does not appear robust ( or is even contradictory ). This becomes apparent by examining the seasonal variation of radiance for the scenarios and again, with respect to a baseline – the PreIndustrial:


While the Double scenario remains consistently more irradiated than the PreIndustrial scenario, and the LGM remains consistently less irradiated ( net ), the HCO and Eemian scenarios indicate great dynamic range, being at times much more or less irradiated for a given latitude. This is consistent with the very strong orbital influence of incoming solar radiance.

Rather than looking at the global annual average of radiance, examining the difference in the annual maximum of net radiance for only the Northern glacial latitudes ( 45N to 83N ), a different pattern emerges:


The source of the glacial and inter-glacial climatic variations would appear to be much more the result of the maximal radiance at latitudes where ice might accumulate or recede than to global average net radiance. This does not contradict global warming. However, it remains an important difference with glacial cycle climates.

A related comparison/contrast is with global temperature. The ice core proxies indicate  temperature anomalies higher for Eemian than HCO, and higher for the HCO than 2010. But the global average radiance above, for these scenarios, indicate lower values for the Eemian and HCO than 2010 ( or PreIndustrial ). Why? Well…

  • The scenarios could be incomplete. The land surface of the Eemian and HCO are thought to have been wetter and more vegetated ( and the LGM direr and more desert ) which would have changed the surface albedo, not reflected here, because I used the monthly CERES calculate surface albedo through out. Of course, the surface albedo is modulated by the cloud albedo which might also have been different during past climates as well.
  • Or, the sea ice of the Eemian and HCO may have been much lower during summer.
  • Or, the temperature proxies don’t represent the global mean. Oxygen isotopes from precipitated snowfall might be skewed toward the season of max precipitation ( summer ). Were the Eeemian and HCO to have greater summer to winter temperature variation, ice core proxies might indicate greater temperatures which were just greater summer temperatures matched by colder winter temperatures.
  • Or, perhaps the global average temperature might have actually increased, but as a consequence of the maximum radiance rather than the annual mean, with some untold seasonal response.


Examining the annual range of net radiance by latitude for the scenarios exhibits:

  • Most of the dynamic range of the glacial scenarios was in the Northern hemisphere
  • The greatest LGM radiance deficit was in the high Northern latitudes
  • The Northern Eemian had the highest high irradiance ( and the lowest lows ).
  • The Northern HCO had the second highest highs ( and the second lowest lows ).
  • The Double CO2 scenario has comparatively small increases in maximum radiance
  • For Double, radiance maximum is greatest at the tropics, least at the poles.
  • The LGM indicates a maximum radiance greater than present in the tropics
  • The annual range of net radiance is greatest at mid-latitudes
  • The annual range of net radiance is comparatively very small at the equator


Jet Stream

Perhaps the most important determinant of atmospheric motion and resultant climate is the meridional gradient of net radiance caused by the nearly spheroidal shape of earth. As mentioned in the last post, the gradient is described as: “Averaged over the year, there is a net energy surplus at the equator and a net energy deficit at the poles. This equator-versus-pole energy imbalance is the fundamental driver of atmospheric and oceanic circulation.”


Deficits at the poles are made up by surplus near the equator.

The gradient of net radiance results in a gradient of temperature, and by the equation of state, a gradient of pressure which results in the wind field associated with the jet stream or storm track. As seen in the seasonal variation animation of radiances above, the location, shape and intensity of the gradient changes through the course of the year.  As a coarse measure, I compared the gradient of the net radiance for each hemisphere by subtracting the spatially weighted net radiance from the pole to 45 degrees from the net radiance of 45 degrees to the equator. The result:


From this, I note:

  • 2010 & Double exhibit little change from Pre Industrial
  • PreIndustrial, 2010 & Double exhibit the strongest winter time gradients
  • LGM exhibits the strongest annual gradient
  • LGM indicates the smallest variation of radiance
  • LGM exhibits maximal gradient from April through October

For the Southern Hemisphere:


  • LGM gradients were also strongest during summer
  • LGM also indicated the smallest range of gradient


Inter-Tropical Convergence Zone

Another important feature of global climate is the Inter Tropical Convergence Zone ( ITCZ ). The importance of both the Jet Streams and the ITCZ can be seen in this idealized global distribution of precipitation:


While the Jet Streams occur because of gradients of upper level pressure formed by gradients of upper level temperature in turn formed by the gradient of net radiance, the ITCZ occurs as a result of motions originating near the surface, not the upper levels. The tropical atmosphere is largely unconditionally stable aloft, meaning that parcels of air aloft, when raised to a higher level, will remain stable and tend to revert to the original level. This is not so for parcels of air near the surface. Parcels from the surface are conditionally unstable. If these parcels are raised to the level at which the water vapour condenses, the parcels will can be warmer than the surrounding atmosphere, and thus tend to rise even more, all the way to the stratosphere.

It is surface polar anticyclones from each pole, arriving near the equator which provide the convergence of the ITCZ. This process can readily be observed by examining global satellite imagery and tracking frontal cloud boundaries converging, and that’s just what what Marcel LeRoux did.These anticyclones are modulated by the continents ( notably the Andes of South America and the higher terrain of Africa ) which explains the main core of the ITCZ lying North of the Equator in the Eastern Pacific as well as the Atlantic. Otherwise, the ITCZ tends to bounce from hemisphere to hemisphere. As a coarse measure of this, I considered the gradient of Northern versus Southern net radiance:GRADIENT_HEMISPHERES

  • Doubling CO2 does not change the inter-hemispheric gradient significantly
  • The seasonal range of gradient was least for the LGM scenario
  • The seasonal range of gradient was greatest for Eemian and HCO

The location and range of the ITCZ would seem to be multifactoral, but range of inter-hemispheric net radiance gradient is consistent with a wider range of monsoonal climate during the Eemian and HCO and a more constrained range of the ITCZ and related tropical precipitation during the LGM. The wetter inter-glacials and drier glacials is indicated by the paleo-vegetation estimates.

No Ice

The scenarios above use the 2010 sea ice for both 2010 and the PreIndustrial scenarios. For the LGM scenario, the NSIDC estimates of sea ice ( moderated by latitude for seasons ) apply. For the Eemian, HCO, Double, and Next Maxima, a 10% reduction from the 2010 sea ice is used for all season. The surface albedo of ice and water differ, of course, so this value can have a large impact. The Eemian and HCO might have had a much greater reduction in sea ice. If so, this reduction might account for some amount of presumed warming of these periods. Also, sea ice decline remains an unknown for the future. With that in mind, I reran the Double scenario with no sea ice whatsoever. The resulting range of net radiances is seen below ( No Sea Ice case is purple ):


And the effect of ice albedo feedback, as difference from PreIndustrial, 2010, and Double with 10% reduction in sea ice, is apparent in the Northern hemisphere:hemispheres:


The effect is also pronounced in the Southern hemisphere:


The difference in radiance appears for the summer months when the ocean surface is exposed to sunshine. In reality, sea ice would not go toward zero during winter and latent heat of freezing would appear while during the melt season, latent heat of melting would come out of the atmosphere. The effect would be to spread the net radiance difference of a ‘no ice’ case over the remaining two thirds of the year, moderating somewhat the temperature impact. Such a temperature response appears not only in recent decades, but also for the early twentieth century warming, presumably also correlated with an Arctic Sea Ice decline:

The effect on inter-hemispheric balance is much smaller, and tends to slightly increase the range of inter-hemispheric imbalance:



A major factor not mentioned while entertaining ideas of jet stream and ITCZ changes is particular to the LGM. Sea level during the LGM was 120 meters lower. In addition to the radiative changes from a thicker atmosphere over the oceans and a thinner atmosphere over the continents, this also increased the effect of continents on circulation. Mountains became 120 meters higher and less passable by dense polar anti-cyclones.  So part of the LGM being drier and less vegetated was in part due to a greater impact by the continents themselves. Also, the impact of the ice sheets is missed by assumption of the 2010 atmosphere. Katabatic winds from the ice were theorized to have been quite intense. An of course, water vapor may well  have changed, though presumably in ways to amplify the above assumed scenarios.

The RF scenarios include the above mentioned uncertainties, assumptions and unknowns, and more, but I believe this exercise is helpful in considering AGW in terms of past fluctuations.

Update ( May 27, 2016 )

Clive Best noted the difference between the 2010 and Double scenarios was less than the nominal 3.7W/m^2 for a CO2 doubling. I neglected to detail that the above radiances are evaluated at the top of the atmosphere (TOA) rather than at the tropopause. Also, no there are no adjustments to the atmosphere including the stratosphere, which is customary when considering the effects of CO2. I intend to duplicate these runs with a CO2 stratospheric temperature adjustment and evaluate at the tropopause in the near future and post the results.

This entry was posted in Uncategorized. Bookmark the permalink.

1 Response to Global Warming in the Context of Glacial Cycles

  1. Pingback: CO2 and the Last Glacial Maximum | Turbulent Eddies

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s