Seasons and Dust Storms on Mars

Two numbers set up everything else. The first is the axial tilt, 25.19 degrees, which is within two degrees of Earth's 23.44. That is the whole reason Mars has seasons, and it means they work the way ours do: each hemisphere leans sunward for half the year and away for the other half, with equinoxes and solstices in between. The second number is the orbital eccentricity, 0.0935, which is five and a half times Earth's 0.0167. Mars swings from 206.65 million kilometres from the Sun at perihelion to 249.261 million at aphelion, and by the inverse square law that is about 45 per cent more sunlight at the near end than the far end. Earth's equivalent variation is about 7 per cent, small enough that most people never notice it. On Mars it dominates. A planet moves fastest when it is closest to the Sun, so the seasons that contain perihelion are compressed and the ones that contain aphelion are stretched. The result, measured in sols, is northern spring 193.3, northern summer 178.6, northern autumn 142.7 and northern winter 154.0, adding to a year of 668.6 sols or 687 Earth days. Read that from the southern hemisphere and it says southern summer is 154 sols long and takes place close to perihelion, while northern summer is 178.6 sols long and happens near aphelion. Southern summer is short, bright and violent. Northern summer is long and mild. Almost everything interesting about Martian weather follows from that asymmetry.

Now the fact that ought to be more famous than it is. Through the long polar winter the pole sits in darkness and radiates heat to space with no sunlight to replace it, and the air above it cools until carbon dioxide stops being a gas. Because the atmosphere is 95 per cent CO2, that does not produce a minor frost. It removes a large fraction of the entire atmosphere from circulation and stacks it on the ground as ice. Leighton and Murray predicted this in 1966, before anyone had measured Martian surface pressure, and the Viking landers confirmed it in the 1970s: daily mean pressure at the Viking 1 site ran between about 6.9 and 9 mbar over one year, and Perseverance watched the same curve at Jezero, mean daily pressure sliding from 761 to 625 pascals. How much of the atmosphere is involved is a genuinely contested number, and the published figures are not measuring the same thing. Schmidt and colleagues put the condensation and sublimation cycle at about one quarter of atmospheric mass. Smith, Zuber and Neumann, reading the caps' thickness with a laser altimeter, say as much as a third over the year. Genova and colleagues, weighing the caps through their pull on orbiting spacecraft, found up to four trillion tonnes on the northern cap alone, about a sixth of the total. A quarter is the usual shorthand. A sixth to a third is the honest range.

Dust gets into the air because something lifts it, and that turns out to be harder to pin down than it sounds. Individual dust grains, finer than 62 micrometres, stick to each other and to the ground, and the thin air struggles to peel them off directly. The efficient route is saltation: wind gets sand grains around 100 micrometres bouncing along the surface, and each impact knocks fine dust into suspension. Older wind tunnel work and models put the threshold at wind speeds of roughly 15 to 22 m/s at 1.5 metres above the ground, which landed missions almost never recorded, yet orbiters kept seeing fresh dune movement and dust devil tracks. That contradiction has a name in the literature, the Martian sand transport puzzle, and it is not fully resolved. A more recent wind tunnel experiment put the threshold far lower still. In-situ peaks are higher than once thought: about 31 m/s at InSight, about 32 m/s at Perseverance. And in October 2025 Bickel and colleagues tracked 1,039 dust devils from orbit and found their drift, and so the ambient wind carrying them, running up to about 44 m/s and routinely exceeding both the saltation threshold and what circulation models predict. Meanwhile the annual pattern is orderly. In years without a global event, three regional storm sequences recur in the same seasonal order, labelled A, B and C, starting near Ls 225, 250 and 310. Dust storms on Mars are not rare weather. They are the climate.

Planet-encircling storms are the rare ones, and 'planet-encircling' is the careful term because the dust encircles the planet rather than obliterating it. Eight are confirmed in the last seventy years, in 1956, 1971, 1973, 1977 twice, 2001, 2007 and 2018, all beginning between Ls 185 and Ls 300, in the perihelion half of the year. Even that count is a matter of which catalogue you use: Richard McKim's telescopic survey for the British Astronomical Association lists ten events since 1873, adding 1909, 1924, 1975 and 1982, and Shirley and colleagues work from a catalogue that includes an 1877 event. The 2018 storm is the best documented. It began on 30 May in Acidalia Planitia as a patch of about 140,000 square kilometres, the first confirmed global storm to start in the northern hemisphere, spread south down the Acidalia corridor at around 5 m/s, crossed the equator, and by 8 June was running east along the southern cap edge at 40 m/s. At Gale crater, Curiosity watched optical depth climb from about 0.6 to 8.5 in twelve sols, ultraviolet sunlight at the ground drop by 97 per cent and the daily temperature swing collapse by 30 K, then decay away over a 43-sol time constant. Streeter and colleagues later found the storm warmed the planet's surface on average, by 0.9 K, because the nightside warming from dust radiating downward outweighed the daytime shading.

The word 'storm' is doing damage here. At six millibars the air has almost no mass to throw at anything. Surface density is about 0.020 kg per cubic metre against 1.217 at Earth's sea level, so Martian air is roughly one sixtieth as dense, and the force a wind exerts scales with density times speed squared. Take the fastest wind ever measured at a Mars landing site, about 32 m/s at Perseverance, which sounds like 72 miles per hour. Run the arithmetic and it pushes about as hard as a 4 m/s breeze on Earth, roughly 9 miles per hour. NASA has said the same thing plainly: even the wind in the largest dust storms likely could not tip or rip apart major mechanical equipment. The opening of The Martian, in which a gale nearly blows a crew off the planet, is good drama and bad physics. What dust storms actually do is block light and stick to things. Opportunity died because opacity over Perseverance Valley hit an estimated 10.8, against a normal 0.5, and its solar arrays could not charge; Spirit and Opportunity both nearly went the same way in 2007. InSight was killed by something slower still, dust accumulating on its panels; NASA's InSight team put the obscuration at about 80 per cent, a figure from mission press material rather than a paper. Curiosity and Perseverance, which run on plutonium rather than sunlight, worked straight through 2018.

What nobody can do is tell you which year the storm will be global. Every Martian year produces regional storms; roughly one year in three or four produces a planet-encircling one, and the trigger is unknown. The standard explanation, a runaway feedback in which dust absorbs sunlight, heats the air, strengthens the Hadley circulation and lifts more dust, works for storms that start near the southern summer solstice but cannot explain the 2001 and 2018 events, which began near the equinox when that circulation is roughly symmetric about the equator. Thermal tides and long-range teleconnections have been proposed instead. Another family of ideas blames the supply side: perhaps the surface dust reservoirs need several years to recharge. The most publicised proposal is James Shirley's orbit-spin coupling, which holds that a term arising from Mars's orbital angular momentum forces the atmosphere directly. It matched the historical catalogue well enough that Shirley, McKim, Battalio and Kass published conditional forecasts for Mars Years 35 through 40 in 2020, and it was built into a general circulation model by Newman and colleagues in 2019. In February 2026 Thomas Pierron and François Forget published a re-derivation of the inertial forces on a parcel of Martian air and concluded that the coupling term comes from an inconsistent application of the transport theorem and should not be in the momentum equation at all, leaving the variability to atmospheric dynamics, radiative feedbacks and dust supply. That argument is open as this is written, in the closing weeks of Mars Year 38, which ends on 30 September 2026.