4.1 The ice ages and the Earth's climate

The Milankovitch theory states that the changes in the Earth's orbit lead to variations in the amount of solar radiation reaching the atmosphere, causing ice ages and warming cycles. There are three types of orbital changes. First, the orbit of the Earth is not circular, but elliptical, and the eccentricity of this ellipse varies over a 100,000 year cycle. Second, there is a tilt or obliquity to the Earth's axis relative to the plane of the solar system, and this cycles over a period of 41,000 years. Third, at any given obliquity, the Earth's axis is wobbling like a top over a 21,000 year cycle and this is called precession.[48]

The Milankovitch theory has been tested and refined by hundreds of investigators since the eponymous Milutin Milankovitch, a Serbian mathematician working during the 1920s.[49] One landmark article (Hays, Imbrie and Shackleton, 1976), compares the Earth's climate history to its orbital history and concludes that the 100,000 year eccentricity cycle has been the predominant driver of the ice ages, although precession and obliquity do explain some climatic variance. The authors do not specify how changes in the delivery of solar radiation to the atmosphere translate into changes in climate: "We avoid the obligation of identifying the physical mechanism of this response...". This is not a failing of the authors; rather it reflects the current limits of scientific knowledge.[50]

The Milankovitch theory explains only a part of the phenomenon we seek to understand. Between an ice age and a warm period there is a 5 degree Celsius difference in the average global temperature. Yet the change in solar energy reaching the Earth due to orbital eccentricity accounts for only a .15 degree Celsius temperature change.[51] The rest of the 4.85 degree C temperature change is thought to be due to climatic feedback effects, but there is no consensus understanding on how these work. [52] Nor do we know what triggers (forces) ice ages. One of the leaders in the field opined in 1998, "We have but simple theories for the ultimate forcing of climate change, even in recent times."[53]

Milkanovitch is the reigning theory because it is the most validated model we have. We can be sure that this small piece of the puzzle is accurate. Since science knows so little about how the ice ages start and terminate, when a new piece of knowledge becomes available, such as MM and IPD, scientists in various fields hope that the new information will resolve long standing dilemmas.

Muller and MacDonald (1995, 1997) argue that for the ice ages of the past 1 million years, the cause may have been MM and IPD.[54] They note that for the past 1 million years there has been a 100 thousand year climate cycle. As stated above, this had been thought to match the eccentricity cycle, but Muller and MacDonald report that the inclination of the orbital plane of the Earth has also been shown to happen in a 100,000 year cycle.[55] This is a fourth characteristic of the Earth's orbit: the plane of the Earth's orbit varies from being perfectly level with the sun.[56] Muller and MacDonald present a series of elegant graphs that match the ice age predictions of the orbital inclination theory to the actual data on the climate cycle, and argue that the inclination theory is the best fit to the climate data:

Figure 1. Column A represents the predicted ice age frequency based on three theories. First is an eccentricity based model, second is a precession based model, and third is Muller and MacDonald's orbital inclination based model. The orbital inclination model clearly and cleanly predicts an ice age every 100,000 years. Moving to columns B,C and D, delta¹⁸O is an isotope that is a proxy for global ice levels. A spike in the delta¹⁸O level indicates an ice age. The orbital inclination model is the best match for the empirical data in these three columns. Science 277 (1997); 215-218, figure 3

What about an increased angle of inclination causes an ice age? Orbital inclination has no significant effect on solar insolation.[57] They hypothesize that the Earth's orbital plane is tilting into a massive stream extraterrestrial dust and that the dust is the trigger mechanism of the ice ages. This theory has received empirical support from a series of analyses of ancient sediment which have shown that there is a 100,000 year periodicity in the accretion rate of MM and IPD, with a sudden increase in the amplitude of the cycle beginning 1 million years ago and continuing today.[58] Muller (1997) concluded, "As far as we know, none of the present climate change models include the effects of dust and meteors. And yet our data suggests that such accretion played the dominant role in the climate for the last million years."[59]

Muller does not identify the precise atmospheric or combined mechanism by which the extraterrestrial dust influences the climate, but Hays, Imbrie and Shackleton (1976) also do not specify a mechanism. K.A. Farley contends that aerosol cooling cannot be the mechanism at work.[60] He maintains that even at the peak of the 100,000 year cycle there is not enough extraterrestrial dust to cause an aerosol cooling effect.[61]

Kortenkamp and Dermott analyze the Muller and MacDonald hypothesis and find against it.[62] They assert that most of the dust probably comes from asteroids and agree that there is a 100,000 year periodicity in dust flux from the asteroid source. However, based on their simulations of dust particle orbits and the Earth's orbit, they conclude that the dust is not concentrated in any particular plane. Instead, the cause of the dust periodicity cycle is indeed the eccentricity of the Earth's orbit. The Earth accumulates up to 3 times more dust during the circular phase of the eccentricity cycle. This is because when traveling in a circular orbit the Earth moves more slowly through the cloud of dust particles, which lowers the velocity of the particles relative to the Earth, making the Earth's gravity more effective at capturing the particles.[63][64]

Kortenkamp and Dermott present their own hypothesis in place of Muller and McDonald's.[65] They note that our current knowledge suggests that many asteroids are clumps of rocky rubble held together by their own weak gravity. If an asteroid of 10 km - 100 km radius was broken apart in a collision with another asteroid, that would "instantaneously liberate," as they put it, a mass of MM and IPD up 10,000 times greater than the normal amount in the solar system. Such an event would occur once every few tens of millions of years. It would lead to about 10¹⁰ kg/year of MM and IDP hitting the stratosphere,[66] the same amount of dust pumped into the stratosphere by a volcanic eruption.[67] This would continue for about 10,000 years as the space dust is gradually captured by Earth's gravity. Since volcanic eruptions are known to cool the climate, 10,000 years of such effects "may lead to substantial changes in the Earth's climate lasting for many millions of years." Additionally, some of the debris would be in the kilometer size asteroid range, which material would arrive in an Earth-crossing orbit towards the latter end of the dust input. The cumulative result of the rubble input could be a gradual mass extinction, the authors note.

Kortenkamp and Dermott's hypothesis would seem to merit investigation by climatologists. [68] Most volcanic aerosols of terrestrial origin settle back to the ground in days or weeks. Only a small fraction reach high enough altitude to remain aloft for extended periods of time. Yet every single particle of asteroidal dust would be injected into the upper atmosphere and so remain airborne for months. The recent Pinatubo eruptions involved far less material and caused a measurable cooling. Even if the dust influx caused cooling of only a few tenths degree, that could be enough of a trigger because the cause-effect relationship is probably non-linear.[69]

One possible causal sequence is that a large spike of asteroidal dust could contribute a significantly greater supply of cloud condensation nuclei, thus increasing the numbers of clouds, which would increase the albedo of the Earth, which could in turn lead to an ice age.[70] Cloud condensation nuclei are small particles from .2 microns up to >1 micron in size which pervade the atmosphere. To be effective, the nuclei must be hygroscopic - water attracting. Water vapor condenses around these particles, leading to cloud formation. Without condensation nuclei, relative humidities of several hundred percent would be required before vapor condensation could begin.[71] Several questions would need to be investigated in order to evaluate this hypothesis. We know the baseline size distribution of MM and IPD (see fn 22 and 23). But we are here dealing with exceptional events that cause a spike in the baseline rate. Do these collisions deliver enough particles at the sizes conducive to cloud formation? Are the particles conducive to the condensation process? The answers might differ with each asteroid collision because different bodies have unique characteristics.

The possible involvement of extraterrestrial matter in glacial cycles suggests that inputs of matter have significantly influenced the atmospheric component of the Earth system, and in turn, the biosphere. The research may have some impact on the debate over global warming. Professor Stephen H. Schneider of Stanford writes, "...we cannot forecast the future accurately without understanding and modeling a great deal of the Earth's past.[72] ... We also need to understand and be able to model the factors that might induce changes in climate, factors we call forcings."[73] Mark H. Thiemens, professor of chemistry says, "One always hears the argument, 'Isn't this all part of a natural cycle?' To answer this question, you really want to have a large scale record."[74] MM and IPD may be necessary components to such an understanding.