MULTIPLE INSTABILITIES AND MODES OF GLACIAL RHYTHMICITY IN THE PLIOPLEISTOCENE - A GENERAL-THEORY OF LATE CENOZOIC CLIMATIC-CHANGE

It has been noted that several distinct modes of glacial oscillation h ave existed during the past few million years, ranging from low-amplit ude, high-frequency oscillations in the early Pliocene, through relati vely high amplitude, predominantly near-40 ky period, oscillations in the late Pliocene and early Pleistocene, to the major near-100 ky peri od oscillations of the late Pleistocene. In addition to other plausibl e mechanisms suggested previously to explain aspects of this multi-rhy thmic phenomenon, we now illustrate another possible contributor to th is type of behavior based on the hypothesis that the slow-response cli matic system is bistable and that two kinds of internal instability ma y be operative along with externally imposed forcing due to earth-orbi tal (Milankovitch) radiation changes and slow, tectonically-induced ch anges in atmospheric carbon dioxide. These two instabilities have been discussed previously: one is due to positive feedback in the global c arbon cycle leading to near-100 ky free oscillations of the ice sheets , and the other is due to the potential for ice-calving catastrophes a ssociated with bedrock variations that can lead to oscillations of a p eriod near 40 ky, independent of obliquity forcing. Within the framewo rk of a dynamical model containing the possibility for these two insta bilities, as well as for stable modes, we show (1) how Milankovitch ra diative changes or stochastic forcing influencing ice sheets can induc e aperiodic (chaotic) transitions between the possible stable and unst able modes, and more significantly, (2) how progressive, long-term, te ctonically-induced, changes in carbon dioxide, acting in concert with earth-orbital radiative variations in high Northern Hemisphere latitud es, can force systematic transitions between the modes. Such systemati c changes can result in an ice mass chronology for the past 5 My that is qualitatively similar to the observed record of global ice mass. In essence, we have constructed a minimum dynamical model of the late Ce nozoic climatic changes, containing what are believed to be the main p hysical factors determiningthese changes: ice mass, bedrock depression , atmospheric ic carbon dioxide concentration, deep ocean thermohaline state, Milankovitch radiation forcing, and slow tectonically-induced carbon dioxide forcing. This model forms the basis for a coherent theo ry for the complex climatic events of this long period.