Science3 publishers3 min readPublished
Roughness maps push Mercury's total contraction up by 10 to 30 percent
Mercury's lumpiest terrain carries the fewest shrinkage wrinkles, which the authors read as impact debris burying the record. Correct for that and total contraction rises to as much as 23 km, with the thermal models still to be redone.
The Scientist · Science desk
What happened
- A study appearing in Geophysical Research Letters on September 10, 2026 concludes Mercury contracted 10% to 30% more than previously thought, with impact-crater debris obscuring the evidence.
- For the first time, whole-surface roughness maps laid over existing maps of contraction features showed that Mercury's roughest areas carry the fewest visible shrinkage wrinkles.
- Correcting for that obscuration puts total diameter loss at up to 23 km over Mercury's lifetime, against a previously estimated range of 4 to 16 km.
- The MESSENGER data behind the estimate resolve only features larger than about 5 km across, and Nishiyama says the revised figures could still be too low.
- Contraction is the observable measure of how much Mercury has cooled, which is what makes its size the input to models of the planet's evolution.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- decision Anyone modelling Mercury's thermal history now has to choose between an observable the authors have superseded and one whose consequences nobody has worked out yet, because the lead author puts that rerun in the future tense.
- contradiction A reader leaves Live Science with 19 km of diameter loss and phys.org's text with up to 23 km, so quoting one figure for Mercury's contraction means inheriting an editorial choice rather than a measurement.
- constraint The correction is bounded by the instrument that produced it: with a 5 km floor on feature size, the band cannot tighten from the existing data, only widen upward.
- precedent Byrne says the moon and Mars likely carry the same bias, so the correction would extend to any cratered body measured only from orbit, and several published contraction budgets for those worlds would then need revisiting.
Roughness is a confounder here, and the paper's move is to treat it as one. Cooling should contract a planet fairly evenly, so shortening structures ought to be about equally common across its surface [20]. Mercury's are not: the lumpiest ground carries the fewest visible ones [4]. Two readings fit that anticorrelation. Rubble from later impacts buried ridges that are there, or rough ground deformed less in the first place. Gaku Nishiyama of the German Aerospace Center Institute of Space Research, the lead author, took the first reading, and his framing is explicitly inferential: the pattern "made us think that there's a process obscuring shortening structures" [3][5]. The planetwide figure then comes from extrapolating the contraction budget of less disrupted terrain into the rough patches [6].
The fraction travels better than the absolute. Apply 10 to 30 percent to the old range of 4 to 16 km and you get 4.4 to 20.8 km [4], while the new upper figure of 23 km sits about 44 percent above the old ceiling [3]. Scientific American sets Paul Byrne's 2013 conference figure of roughly 11 km against that 23 km [18], and neither account states whether a given number describes a radius or a diameter. The distinction is a factor of two: 11 km of radius is 22 km of diameter [5]. Until the convention is pinned down, the percentage is the sturdier number.
That total is a magnitude, not a rate. Mercury has been shedding the heat of its formation for something like 4.5 billion years [20], and 23 km against a present diameter of 4,880 km is 0.47 percent of the planet [9][2]. Fit to cooling physics is what makes this number interesting, rather than its size on its own. According to phys.org, the higher estimate brings observed contraction more into step with what cooling physics predicts [14], and more contraction implies a larger metal core, fewer light elements such as silicon mixed into it, or a hotter starting temperature [16]. None of those scenarios has been recomputed. Nishiyama told Scientific American that the new estimate "would tell a different scenario of Mercury's thermal evolution," and called working out that implication "the next question that we have to tackle" [15].
Byrne, who was not involved in the work, told Scientific American he is confident the results will hold [17]. That confidence covers the direction of the correction rather than its size, and size is what a cooling model consumes. Nishiyama calls 30 percent "a little bit surprising" while accepting the corrected figure [23]. One spacecraft has ever entered orbit around Mercury [21], and the maps behind this estimate come from it [10]. Read the result as a floor that will move when the resolution does.
What to watch
- Whether BepiColombo's higher-resolution scans, starting November 2026, find small scarps and ridges concentrated in the rough terrain the obscuration argument predicts [c12].
- Whether the laser altimetry Nishiyama points to raises the contraction estimate again or holds it where it is [c13].
- Whether anyone applies the same roughness correction to the moon and Mars, as Byrne urges [c19].