226
A. Paul· W. H. Berger
would yield a split peak in the output signal. Hence, if there were an unsplit
peak with a period near 100 ka in the response of such a climate system, then
eccentricity could not be the driving force, but orbital inclination could.
Muller and MacDonald not only argue that the 100-ka glacial cycle is based
on changes in orbital inclination, they also propose that the mid-Pleistocene
climate transition might have been caused by an increase in the amount of
meteoroids or dust (Muller 1994; Muller and MacDonald 1997a). In a study of
helium-3 in oceanic sediments, Farley (1995) indeed reported a dramatic rise
of the IDP accretion rate at about 1 Ma B. P. Since the coincident onset of both
the major 100-ka glacial cycle and a higher IDP flux could hardly be fortuitous
and could not be attributed to the regular changes in insolation, it might indicate a causal relationship.
There are a few additional pieces of evidence that support the inclinationaccretion hypothesis. Farley and Patterson (1995) detected a 100-ka cycle of
helium-3 in oceanic sediments (where the primary source for this isotope is
accretion of IDP). A narrow ring of IDP extending only 2° from the invariable
plane was observed by the Infrared Astronomical Satellite (IRAS; Sykes 1988).
The number of noctilucent clouds (mesospheric clouds associated with the effects of high meteors and high-altitude dust) shows a strong peak within about
a day of the date when the Earth passes through the invariable plane, supporting the contention that accretion increases significantly at such a date (Foyle
and Haurwitz 1966; Thomas 1991).
We add that Farley and Patterson themselves consider the possibility that
100-ka climate effects could control the observed helium-3 flux, by focusing
of IDP to particular regions of the seafloor; but they prefer an interpretation
in which the 100-ka cycle in IDP accretion is the cause of helium-3 variability.
In a subsequent study, Marcantonio et al. (1996) compare 3He and 230Th in two
deep-sea cores from the equatorial Pacific Ocean and conclude that the helium3 variability is indeed due to IDP focusing. They find a temporal pattern that
points to a causal relationship between climate cycles and cycles of reorganization of near-bottom oceanic currents.
3
The Paleoclimate Model
As described in detail in Saltzman and Verbitsky (1992; 1993), the "full PCM"
is composed of four time-dependent equations governing slow-response variables and three diagnostic equations for fast-response variables. The slow-response variables are total ice mass of the Northern Hemisphere ice sheets excluding the Greenland ice sheet \11, mean bedrock depression below the undepressed topography D (due to each of n individual ice sheets of assumed equal
mass \11 In), atmospheric carbon dioxide concentration t-t and mean ocean temperature fJ. The fast-response variables are summer mean temperature at high
latitudes of the Northern Hemisphere T, mean elevation of the undepressed
bedrock above sea level Z and an ice-calving catastrophe function C, adapted
A. Paul· W. H. Berger
would yield a split peak in the output signal. Hence, if there were an unsplit
peak with a period near 100 ka in the response of such a climate system, then
eccentricity could not be the driving force, but orbital inclination could.
Muller and MacDonald not only argue that the 100-ka glacial cycle is based
on changes in orbital inclination, they also propose that the mid-Pleistocene
climate transition might have been caused by an increase in the amount of
meteoroids or dust (Muller 1994; Muller and MacDonald 1997a). In a study of
helium-3 in oceanic sediments, Farley (1995) indeed reported a dramatic rise
of the IDP accretion rate at about 1 Ma B. P. Since the coincident onset of both
the major 100-ka glacial cycle and a higher IDP flux could hardly be fortuitous
and could not be attributed to the regular changes in insolation, it might indicate a causal relationship.
There are a few additional pieces of evidence that support the inclinationaccretion hypothesis. Farley and Patterson (1995) detected a 100-ka cycle of
helium-3 in oceanic sediments (where the primary source for this isotope is
accretion of IDP). A narrow ring of IDP extending only 2° from the invariable
plane was observed by the Infrared Astronomical Satellite (IRAS; Sykes 1988).
The number of noctilucent clouds (mesospheric clouds associated with the effects of high meteors and high-altitude dust) shows a strong peak within about
a day of the date when the Earth passes through the invariable plane, supporting the contention that accretion increases significantly at such a date (Foyle
and Haurwitz 1966; Thomas 1991).
We add that Farley and Patterson themselves consider the possibility that
100-ka climate effects could control the observed helium-3 flux, by focusing
of IDP to particular regions of the seafloor; but they prefer an interpretation
in which the 100-ka cycle in IDP accretion is the cause of helium-3 variability.
In a subsequent study, Marcantonio et al. (1996) compare 3He and 230Th in two
deep-sea cores from the equatorial Pacific Ocean and conclude that the helium3 variability is indeed due to IDP focusing. They find a temporal pattern that
points to a causal relationship between climate cycles and cycles of reorganization of near-bottom oceanic currents.
3
The Paleoclimate Model
As described in detail in Saltzman and Verbitsky (1992; 1993), the "full PCM"
is composed of four time-dependent equations governing slow-response variables and three diagnostic equations for fast-response variables. The slow-response variables are total ice mass of the Northern Hemisphere ice sheets excluding the Greenland ice sheet \11, mean bedrock depression below the undepressed topography D (due to each of n individual ice sheets of assumed equal
mass \11 In), atmospheric carbon dioxide concentration t-t and mean ocean temperature fJ. The fast-response variables are summer mean temperature at high
latitudes of the Northern Hemisphere T, mean elevation of the undepressed
bedrock above sea level Z and an ice-calving catastrophe function C, adapted