The Earths long-term climate changes and ice ages: a derivation of Milankovitch cycles from first principles

TL;DR

This paper derives Earth's long-term climate cycles, known as Milankovitch cycles, from first principles, identifying their main periods and confirming their consistency with previous numerical simulations.

Contribution

It presents an analytical derivation of Milankovitch cycles from planetary precession principles, avoiding numerical simulation issues.

Findings

Main cycle period of 41,000 years due to Earth-Venus precession

Secondary cycle of 29,500 years from Earth-Venus precession modes

Third cycle of 54,000 years influenced by Jupiter and Saturn precession

Abstract

Long-term changes in the tilt of the Earths axis, relative to the plane of its orbit, are of great significance to long-term climate change, because they control the size of the arctic and antarctic circles. These Milankovitch cycles have generally been calculated by numerical integration of Newtons equations of motion, and there is some controversy over the results because they are sensitive to numerical drift over the very long computer simulations involved. In this paper the cycles are calculated from first principles, without any reliance on computer simulation. The problem is one of planetary precession, and is solvable by the methods used to study the precession of a spinning top. It is shown that the main component of Milankovitch cycles has a period of 41,000 years and is due to one of the modes of precession of the Earth-Venus system. The other mode of this system produces a component of period 29,500 years, and a third component of period 54,000 years results from the influence of the precession of the orbits of Jupiter and Saturn. These results agree closely with several of the numerical simulations in the literature, and strongly suggest that other different results in the literature are incorrect.