Mars's tilt turns out to be chaotic
Two 1993 calculations showed the Martian axis wanders chaotically — one integration ran it from 11 to 49 degrees over 80 million years. The present 25.19 degrees is a low value, not a constant.
Two papers eight days apart said the same unwelcome thing. On 18 February 1993 Nature ran Jacques Laskar and Philippe Robutel on the chaotic obliquity of the planets; on 26 February Science ran Jihad Touma and Jack Wisdom's direct numerical integration of the rotation of Mars. The work was independent — Touma and Wisdom had submitted theirs in December 1992. Both concluded that the tilt of the Martian axis is not a number but a trajectory in a chaotic system, and that it cannot be run backwards more than a few tens of millions of years.
That the tilt moved was not new. William Ward had shown in Science in 1973 that it oscillates, and a 1991 paper by Ward and Rudy had it reaching 45 degrees. The character of the motion was new. Mars has no large moon; its spin axis precesses at about 8.26 arcseconds a year, close enough to one of the solar system's secular inclination frequencies — the one that governs the tilt of Venus's orbit plane — that the resonance gets crossed repeatedly instead of cleanly entered. Touma and Wisdom's integration wandered between roughly 11 and 49 degrees over 80 million years, with nearby starting conditions diverging in three to four million years. A companion paper in the same Nature issue made the contrast plain: Earth's tilt holds inside about 22.1 to 24.5 degrees only because the Moon is there.
Laskar's group put numbers on the long run in 2004. A precise solution is possible over 10 to 20 million years and no further; past that only statistics survive. Across five billion years the mean obliquity works out at 37.62 degrees, with a standard deviation of 13.82 and a maximum of 82. Their standard model takes 41.80 degrees as the most probable value. The present 25.19 degrees is a low outlier.
The fingerprint turned up in Mars Global Surveyor data. On 18 December 2003 James Head and colleagues reported a dusty, water-ice-rich mantle, metres thick and layered, draped symmetrically in both hemispheres from about 30 degrees latitude to the poles. They dated its emplacement to roughly 2.1 to 0.4 million years ago, when obliquity reached 30 to 35 degrees, and read its present pitted, retreating state as decay. Mars is in an interglacial. Its ice ages run the wrong way round from ours: warmer poles, ice carried outward toward the equator.
None of this is settled. Holo, Kite and Robbins argued in 2018, from the orientations of elliptical craters, that the mean obliquity since the late Hesperian was probably low — 10 to 30 degrees, with under a fifth of the time spent above 40. What is no longer arguable is the framing. Every measurement of Martian climate we have is one phase of a cycle nobody can rewind.