Abstract. Based upon research results over the past five decades, there has been a general acceptance that the ice ages were initiated by astronomical phenomenon. Specifically, marine, ice and terrestrial paleoclimate data have supported elements of the Milankovitch astronomical theory of the ice ages. However, there remain unresolved problems between the empirical findings and theory. The 100 thousand year problem has been the subject of extensive research since a 100 thousand year cycle that matches the Earth orbit eccentricity period dominates the frequencies found in paleoclimate records. Yet, eccentricity produces an insignificant variation in annual solar energy. Other problems include the Stage 11 problem, the missing interglacials problem, how glaciation is sustained over multiple tens of thousands of years and synchronous hemispheric glaciation. I shall show these problems are resolved by modification of the prevailing Milankovitch theory. In particular, two elements of the theory need modification. One is the limitation of eccentricity's role and the other assuming that glaciation results only from cool summer conditions. By applying the Solar Energy Invariance law to define e-seasons, how eccentricity provides conditions for glaciation is demonstrated. The results show eccentricity variations provide significant solar energy variations at the top of the earth's atmosphere to produce glaciation that is global. Global glaciation results in colder winter glaciation occurring in one hemisphere simultaneous with cool summer glaciation in the other hemisphere. Analysis with these modifications resolves each of the problems.
I.I. Smul’skii. New astronomical theory of the ice ages