Venus time

Two clocks run on Venus, and much of the confusion about its day comes from mixing them up. One is the sidereal day, the time the planet takes to turn once relative to the distant stars. The rotation model adopted by the International Astronomical Union's working group on cartographic coordinates, and distributed to mission software by NASA's NAIF team in its pck00011 constants file, writes the angle of the Venus prime meridian as 160.20 minus 1.4813688 degrees for every day counted from noon on 1 January 2000. That rate is one turn every 243.0185 days. The minus sign is the backward spin, since the angle shrinks with time. NAIF's file records that the Venus values were left unchanged in the working group's 2015 report, so the figure in force today is the one derived from Magellan in the early 1990s.

Backwards needs a reference. NAIF's documentation fixes the north pole of every planet on the north side of the solar system's invariable plane, "regardless of the body's spin direction". Seen from above that pole, Venus turns clockwise, opposite to Earth and opposite to its own path round the Sun. Pick the pole by the right-hand rule instead and the same axis points down, which is why NSSDCA's fact sheet gives Venus an obliquity to its orbit of 177.36 degrees and an inclination of the equator of 2.64 degrees. Both are correct: one tilt measured from opposite ends of the axis. Longitudes follow the IAU pole. The USGS gazetteer lists every Venus feature in planetocentric coordinates, east positive from 0 to 360, and zero longitude runs through the central peak of Ariadne, a crater 23.6 km across at 43.9 degrees north.

The other clock is the solar day, from one sunrise to the next at the same spot. On Earth the orbit makes the solar day a little longer than the sidereal one, 24 hours against 23 hours 56 minutes, because the planet must turn slightly further to face the Sun again. On Venus the orbit runs one way and the spin the other, so the two motions add and the solar day comes out far shorter than the turn: one divided by 116.75 equals one divided by 243.02 plus one divided by 224.70. NSSDCA gives the length of day as 2,802.0 hours, or 116.75 Earth days. The IAU rate and NSSDCA's orbital period of 224.701 days give 116.7505 days by that formula; JPL's Horizons sheet prints 116.7490, two minutes less, without saying why. JAXA's public summary rounds all three: a spin of 243 days, an orbit of 225 and a day of 117.

Here the famous line about a day longer than a year needs care. A turn of 243 days is longer than an orbit of 224.7 days, so Venus does rotate once in more than a year. The day anyone standing there would count, sunrise to sunrise, fits into the year 1.92 times, so noon and midnight each come round a little less than twice per Venus year. At the equator daylight lasts about 58.4 Earth days, half the solar day, and the night is as long. With the axis within 2.64 degrees of upright, day length hardly changes through the year, and JAXA's summary says plainly that Venus has no seasons. NSSDCA lists the day to night temperature range at the surface as roughly zero, so the long night is as hot as the day.

Because the ground turns westward, everything in the sky appears to travel the other way: stars and Sun rise in the west and set in the east. The point beneath the Sun slides eastward over the surface by about 3.08 degrees of longitude per Earth day, so by azmth's arithmetic the Sun needs nearly six Earth hours to move its own width across the sky, about 0.74 degrees as seen from Venus. Soviet landers measured the light at the ground, and NSSDCA's summary of the preliminary Venera 9 results compares it to the light at Earth's midlatitudes on a cloudy summer day. The direction of the Sun's daily motion, west to east, is geometry and holds whether or not its disc can be made out through the overcast.

Clock time at a place on Venus is kept the way Soviet and American teams kept it, as local solar time. The solar day is divided into 24 local hours, with noon when the Sun is overhead and six o'clock at the morning terminator, so each Venus hour lasts 116.75 Earth hours, just under five Earth days. Because the Sun moves east across the ground, a site east of the point beneath the Sun is still in its morning: local time is 12 hours plus the difference between the subsolar longitude and the site's longitude, divided by 15, in east longitudes. Ralph Lorenz worked with exactly that difference in a 2018 note added to the Planetary Data System archive of the Vega balloons, which drifted westward far faster than the Sun moved: the Vega 1 balloon started just after local midnight and Vega 2 at about 1 a.m., and both were last heard at about 8 a.m. local solar time, two Earth days later.

Mission records use the same clock. Venera 4 descended on the night side where, in Marov's account, the local solar time was 4:40 with the Sun 110 degrees from the zenith. Venera 8 landed at 6:24 local solar time with the Sun 84.5 degrees from the zenith, about five and a half degrees above the horizon, according to NSSDCA, which adds that its event times are spacecraft times, the Earth-received time minus 3 minutes 36 seconds of light travel. In the same account, Venera 9 touched down at 13:12 by Venus solar time with the Sun 33 degrees from the zenith. The four Pioneer Venus probes of 9 December 1978 reached the ground within 12 minutes of one another at local times from 0:07 to 7:38, because they were spread across the night and morning hemispheres. No source read for this page used a running count of Venus days like the Mars Sol Date: events are dated in Earth time and placed in the Venus day by local solar time.

Until radar, the clouds kept the length of the turn out of reach. Charles Boyer summed up the telescope record in the French journal L'Astronomie in 1965: every value from 24 hours to 225 days had been put forward and none confirmed, and the generally accepted opinion was that the planet always showed the same face to the Sun. Radar cut through. In Andrew Butrica's NASA history of planetary radar, JPL and MIT's Lincoln Laboratory at first concluded from their 1961 echoes that the day was about as long as the year, around 225 days, with one face always turned to Earth at closest approach. Between 1 October and 17 December 1962, near-daily observations at Goldstone let JPL's Roland Carpenter follow one irregularity in the echo spectrum that persisted from day to day and moved slowly. He took it for a rough region of ground carried round by the rotation and concluded that the sidereal period was 250 plus or minus 40 days, retrograde. He and fellow JPL radar astronomer Dick Goldstein published the result in Science on 8 March 1963.

Others came close. At Lincoln Laboratory, William Smith had seen in 1961 data that only retrograde motion explained how the echo bandwidth changed, but his supervisor Paul Green, wary after a false radar detection of Venus in 1958, had him tone the paper down, and its abstract spoke only of "a very slow or possibly retrograde rotation". Vladimir Kotelnikov's Soviet group first estimated about 11 days from its 1961 echoes and, after the 1962 approach, concluded 200 to 300 days, retrograde, by Butrica's account.

Meanwhile two French observers had found something moving much faster. Henri Camichel had been photographing Venus in ultraviolet light at the Pic du Midi observatory since 1953, and on his advice Charles Boyer, a magistrate who was president of the Court of Appeal in Brazzaville according to Dollfus's 1988 account, began in August 1957 with a 26 cm telescope and a Wratten 34 violet filter. A dark band seen on 28 August was back on 1 September. Boyer wrote that he immediately thought of a rotation in four days and told Camichel, who found the same recurrence in his 1953 plates. The joint summary of their 1961 paper in the Annales d'Astrophysique was cautious: the changing markings could be explained by a displacement in the retrograde direction, similar aspects returned roughly every four days, and this might be the rotation period of Venus, though they could not affirm it. Inside the paper, the authors parted ways. Boyer's section, timing a Y-shaped marking, concluded that Venus turned retrograde with a synodic period very close to 96 hours 33 minutes, a sidereal day of 97 hours 38 minutes, and took the dark masses for clouds that roughly followed features on the ground. Camichel's section placed the photographed details in the atmosphere of the planet rather than on its ground and allowed that they might be purely atmospheric.

Boyer kept measuring. In 1962 an international campaign organised through Dollfus, then president of IAU Commission 16, had eight observatories spread round the globe photograph Venus in ultraviolet through a whole evening elongation. By 1965 Boyer faced radar periods of about 250 days. Citing hour-to-hour motion of the markings, stereo measurements by France's IGN mapping agency and a 1964 Doppler measurement by Guinot of the Paris Observatory, he wrote that the problem could be considered solved and that the planet turned retrograde in about four days. The direction and the four days were right, but what turns in four days is the upper atmosphere.

JAXA describes it this way: the atmosphere of Venus turns in the direction of the spin, westward, and at the top of the cloud layer it circles the planet in only about four days, some 60 times faster than the ground. Analysis of Akatsuki images and temperatures by Takeshi Horinouchi's group, published in Science and announced by JAXA on 24 April 2020, found that thermal tides, waves driven by daytime heating, carry the angular momentum that keeps the fastest cloud-top flow going near the equator. NASA's caption for a reprocessed Mariner 10 view gives the clouds about four and a half days per circuit at about 100 m/s. Venus therefore keeps a third clock, the four-day sky described on the atmosphere page, and it was that clock Boyer and Camichel had been reading.

The number in today's model comes from Magellan's radar images. M. E. Davies and eight co-authors, E. L. Akim among them, followed hundreds of landmarks through Magellan's early mapping and published a period of 243.0185 plus or minus 0.0001 days in August 1992, an average over about 500 days in the description of a later team. The quoted precision, under nine seconds, described those 500 days and could not promise that the planet would keep that rate.

Evidence of drift came from Venus Express. Its VIRTIS spectrometer could make out surface features by peering through the clouds in the infrared, and when Nils Müller of the German Aerospace Center (DLR) lined its maps up with Magellan's, taken some 16 years earlier, features sat up to 20 km from where the Magellan rate put them. "When the two maps did not align, I first thought there was a mistake in my calculations," he told ESA in February 2012. Matching the maps required a day on average 6.5 minutes longer, 243.023 plus or minus 0.001 days in the paper he led in Icarus. ESA noted that this agreed with long-duration radar from Earth, and in 2019 Bruce Campbell and colleagues put 29 years of Earth-based radar images, 1988 to 2017, at 243.0212 plus or minus 0.0006 days. The team weighed short random swings, which should average out, against decades-long weather cycles, which might not; project scientist Håkan Svedhem added that choosing landing sites would need an accurate rate.

Jean-Luc Margot of UCLA and colleagues then measured the spin directly instead of mapping it. From 2006 to 2020 they sent a 3.5 cm radar signal of 200 to 400 kW to Venus from DSS-14, the 70 m antenna at Goldstone, and recorded the echoes there and at the 100 m Green Bank Telescope in West Virginia. The speckled echo from rough ground sweeps across Earth as the planet turns; when its track lines up with the two dishes, the same flicker reaches Green Bank about 20 seconds after Goldstone, and the lag gives the spin rate at that moment. Of 121 attempts, 21 succeeded. Published in Nature Astronomy on 29 April 2021, the results put the average sidereal day for 2006 to 2020 at 243.0226 plus or minus 0.0013 days, while single measurements ranged from 243.0159 to 243.0308 days, a spread of 61 parts per million or about 21 minutes. The team rejected a constant spin with high confidence and found the earlier averages at odds with most of their instantaneous values. That, Margot told UCLA, "probably explains why previous estimates didn't agree with one another".

The likely cause is planetary weather. The atmosphere of Venus carries an estimated angular momentum about 180 times Earth's, and its share of the whole planet's angular momentum is some 60,000 times larger than on Earth. Margot's team calculated that at least about 3.8 percent of it must be moving to and from the solid planet, and concluded that core torques and tides are far too weak to explain the changes. The length-of-day variations they measured are, in their words, three orders of magnitude larger than on Earth, where the day changes by milliseconds. What drives them in detail is open. The data hint at a forcing twice or once per Venus day that the team could not confirm, no long-term trend was detectable, and the atmospheric models available then predicted exchanges 30 to 300 times slower than observed.

For maps and clocks the differences are small in each turn and large over years. Margot's group calculated that the positions of equatorial landmarks predicted by their model and by the Magellan-based model in use drift apart by about 1 km per year, and that the random wandering alone adds 3.3 km of uncertainty over a decade; UCLA's release put the risk for a future landing at up to 30 km. By azmth's arithmetic the two rates have separated by about a quarter of a degree of longitude since the start of 2000, some 26 km at the equator. On a local solar clock a quarter of a degree is about one Venus minute, just under two Earth hours, so a sunrise computed with either rate falls at nearly the same moment. The maps are where it matters, as ESA and UCLA both said of future landers.

Why Venus spins backwards at all has no settled answer. Tides raised by the Sun in the solid planet would on their own lock its spin to its 225-day orbit, Margot's team noted, and Thomas Gold and Steven Soter proposed in 1969 that the Sun's pull on a heat-driven tide in the thick atmosphere pushes the other way, so that the spin settles where the two torques balance. Alexandre Correia and Jacques Laskar showed in Nature in 2001 that a planet with a dense atmosphere ends in one of only four rotation states, and that most starting conditions lead to Venus's present one by two routes: the axis flipping over, or a forward spin slowing and reversing while the tilt goes to zero. Giant impact simulations by Mirco Bussmann and colleagues (Astronomy and Astrophysics, 2025) found many collisions consistent with the present rotation, including head-on hits on a non-rotating Venus, which typically left small debris discs likely to fall back onto the planet instead of forming a lasting moon. In a study its authors list as accepted by The Astrophysical Journal (arXiv version of 21 September 2026), Stephen Kane and colleagues found that tides could both remove an early moon and slow a fast-spinning Venus, though only for a narrow range of starting spins and moon masses.

One near coincidence has tempted astronomers since the 1960s. Venus and Earth line up at inferior conjunction every 583.92 days, which is 5.0015 Venus solar days, so Venus turns almost the same face towards Earth at each close approach. The early radar teams read the pattern as a lock, and Peter Goldreich and Stanton Peale asked in Nature in 1966 whether the spin was resonant with Earth. An exact lock would need a period of about 243.16 days by azmth's arithmetic, against the 243.02 measured. Margot's team concluded that the atmospheric swings are so large that they likely prevent capture in such a resonance.