The astronomical theory and data compared - Ice Ages and Interglacials: Measurements, Interpretation, and Models - page 288

studies point to the need to invoke internal dynamics of ice sheets as a mechanism
for occasional rapid collapses if a threshold size is exceeded. Nor do the results
explain the mid-Pleistocene transition between an earlier interval characterized
by 40 kyr durations of ice ages and a later interval with 80 kyr to 120 kyr
durations ...

. The prevailing view to date has been that ice sheet volume is
the most important variable to consider. While this is obviously the case for
global sea level, it is ice sheet extent that matters most for albedo, and ice sheet
height that matters for atmospheric circulation . Ice sheets are dynamic

systems and these properties can vary quite differently from each other. However,
the results presented here demonstrate the critical physical importance of focus-
ing on the rate of change of ice volume, as opposed to the ice volume itself. The
available evidence supports the essence of the original idea

...

[that] (1) the strong
expectation on physical grounds that summertime insolation is the key player in
the mass balance of great Northern Hemisphere continental ice sheets of the ice
ages; and (2) the rate of change of global ice volume is in antiphase with
variations in summertime insolation in the northern high latitudes that, in turn,
are due to the changing orbit of the Earth.''

...

10.2.4 Terminations of ice ages and origins of interglacials

We have previously seen that ice ages tend to end rather abruptly compared with
their slow rate of formation over many tens of thousands of years. A number of
studies have been carried out in an attempt to understand why this occurs, but
none of these are entirely satisfactory.
Rothlisberger et al. (2008) investigated the phasing between a South American
dust proxy (non sea salt calcium flux, nssCa), a sea ice proxy (sea salt sodium flux,
ssNa), and a proxy for Antarctic temperature (deuterium, d D) to determine
whether a similar sequence of events applied to the last nine ice age terminations.
The connections between nssCa, ssNa, and d D are discussed in the paper; these
did not appear very illuminating to this writer. However, the d D over nine
terminations was very interesting. The data were analyzed using a simple model of
a linear ramp connecting pre- and post-termination properties, with the ramp
constituting the termination period. In some cases, the transition from pre- to
post-termination conditions occurred not merely with one ramp—but several. The
ramp data are summarized in Table 10.3 .
It is interesting to compare the termination periods from Table 10.3 with the
peak solar intensity at 65 N (as shown in Figure 10.9 ). It is rather impressive that
every single termination occurs on an upswing in solar input. However, some of
these upswings in solar input are very modest. The last four terminations were
coincident with stronger upswings in peak solar intensity at 65 N.
We can compare the results of Rothlisberger et al. (2008) for the onset of
terminations with dates compiled by others (as shown in Table 10.4 ). There is
fairly good agreement between investigators.