When a powerful earthquake and tsunami destroyed large parts of Indonesia on 26 December 2004, Earth’s crust was altered sufficiently to change the speed at which the planet rotates. Ever since, days have been 2.68 microseconds shorter.

But the rotations of other planets in the Solar System have been harder to determine. Saturn’s, for instance, is still subject to vast uncertainties.

In Nature this week, three Israeli planetary scientists suggest that a day on Saturn is 15 minutes shorter than previously thought. It lasts for 10 hours, 32 minutes and 45 seconds, they say (R.Helled,E.GalantiandY.KaspiNaturehttp://dx.doi.org/10.1038/nature14278;2015).

Because it is mostly a dense fluid of helium and hydrogen, Saturn would not be expected to have a single, well-defined rate of spin. But the planet is thought to have a rocky core, and to rotate roughly in unison in the same way that a solid object would. (The fluid Sun, by contrast, rotates 34% faster at its equator than at its poles.) Saturn’s hazy atmosphere — the thickness of which is hard to pin down — makes it difficult to estimate how fast any firmer mass inside is spinning.

The two Voyager probes gave us the first modern estimate of Saturn’s rotation period — around 10 hours and 39 minutes — some 35 years ago. They used radiofrequency emissions sent out by the planet’s wobbling magnetic field. NASA’s Cassini spacecraft, which has been orbiting Saturn since 2004, repeated those radio measurements but found a value that was not only substantially larger, but also changed with time. Experts began to question whether the technique, which had been applied successfully to Jupiter, was reliable for its smaller sibling Saturn.

Ravit Helled of Tel Aviv University and her colleagues took an approach that is, in a way, more traditional: they estimated the rotation indirectly, from the way it distorts the planet. Centrifugal forces give spinning celestial bodies a flattened shape, wider at the equator than at the poles. For Saturn, the effect is accentuated by the planet’s size — centrifugal forces are stronger the farther one moves from the axis of rotation — and rapid spin. Moreover, its mostly fluid nature means that the planet deforms more easily. Not much is known about the distribution of mass inside Saturn, but something can be gleaned from its gravitational pull on objects that orbit it. Small deviations in Cassini’s trajectories, for instance, have revealed that the gravitational field is not symmetrical.

This information is still not sufficient to nail down the internal structure of the planet or to calculate the rotation rate. So Helled and her co-authors did the next best thing. They produced a series of likely internal structures that narrow down the uncertainty to a 92-second range.

Are they right? NASA plans to probe Saturn’s gravitational field much more precisely towards the end of the Cassini mission, when the craft will dip into an elongated orbit and fly between the atmosphere and the system of rings. Just before its fuel runs out in 2017, Cassini will perform a controlled plunge into the planet. And the data it gathers along the way should help not just to elucidate the planet’s make-up, but also to test models of how gas giants form in distant star systems. But Saturn will not be alone. From 2016, Juno, a major NASA mission that launched in 2011, will take similar measurements of the gravitational field of Jupiter. Sooner rather than later, it seems, these two gas giants will reveal their mysteries.