Orbital insolation, ice volume, and greenhouse gases

The SPECMAP models of orbital-scale climate change (Imbrie et al., Paleoceanography 7 (1992) 701, Paleoceanography 8 (1993) 699) are the most comprehensive to date: all major climatic observations were analyzed within the framework of the three orbital signals. Subsequently, tuning of signals in Vostok ice to insolation forcing has fixed the timing of greenhouse-gas changes closely enough to permit an assessment of their orbital-scale climatic role. In addition, evidence from several sources has suggested changes in the SPECMAP delta(18)O time scale. This new information indicates that the timing of CO2 changes at the periods of precession and obliquity does not fit the 1992 SPECMAP model of a train of responses initiated in the north, propagated to the south, and later returning north to force the ice sheets. In addition, analysis of the effects of rectification on 100,000-year climatic signals reveals that all have a phase on or near that of eccentricity. This close clustering of phases rules out the long time constants for 100,000-year ice sheets required by the 1993 SPECMAP model. A new hypotheses presented here revives elements of an earlier CLIMAP view (Hays et al., Science 194 (1976a) 1121) but adds a new assessment of the role of greenhouse gases. As proposed by Milankovitch, summer (mid-July) insolation forces northern hemisphere ice sheets at the obliquity and precession periods, with an ice time constant derived here of 10,000 years. Changes in ice volume at 41,000 years drive ice-proximal signals (SST, NADW, dust) that produce a strong positive CO2 feedback and further amplify ice-volume changes. At the precession period, July insolation forces ice sheets but it also drives fast and early responses in CH4 through changes in tropical monsoons and boreal wetlands, and variations in CO2 through southern hemisphere processes. These CH4 and CO2 responses enhance insolation forcing of ice volume. Climatic responses at 100,000 years result from eccentricity pacing of forced processes embedded in obliquity and precession cycles. Increased modulation of precession by eccentricity every 100,000 years produces 23,000-year CO2 and CH4 maxima that enhance ablation caused by summer insolation and drive climate deeper into an interglacial state. When eccentricity modulation decreases at the 100,000-year cycle, ice sheets grow larger in response to obliquity forcing and activate a 41,000-year CO2 feedback that drives climate deeper into a glacial state. Alternation of these forced processes because of eccentricity pacing produces the 1 00,000-year cycle. The 100,000-year cycle began 0.9 Myr ago because gradual global cooling allowed ice sheets to survive during weak precession insolation maxima and grow large enough during 41,000-year ice-volume maxima to generate strong positive CO2 feedback. The natural orbital-scale timing of these processes indicates that ice sheets should have appeared 6000-3500 years ago and that CO, and CH4 concentrations should have fallen steadily from 11,000 years ago until now. But new ice did not appear, and CO2 and CH4 began anomalous increases at 8000 and 5000 years ago, respectively. Human generation of CO2 and CH4 is implicated in these anomalous trends and in the failure of ice sheets to appear in Canada. (C) 2003 Elsevier Science Ltd. All rights reserved.