Three Strikes Against Uranus

Uranus has an axial tilt of 98 degrees, giving it the appearance of a bulls-eye as it revolves around the sun.  Its moons revolve comfortably around the planet’s equator.  This unusual arrangement, unique in the solar system, has challenged planetary scientists since its discovery.  A new model accounts for it through a series of gentle bumps from impacts as the planet was forming from dust and gas, but how would one ever test such an idea?

Simplistic models of planet formation from an evolving dust disk around a star should end up with all the planets revolving in near-circular ellipses in the equatorial plane, and rotating with axes perpendicular to the plane.  That contradicts a great deal of observational evidence; the planets have varying degrees of orbital eccentricity, orbital inclination, and axial tilt.  Some of these can be explained by subsequent gravitational interactions.  In the last decade or so, based on observations of “hot Jupiters” around other stars, radial migration has been invoked to get gas giants to form in the distant recesses of the dust disk, then move them closer to the star just in time (8/21/2009).  Uranus, in particular, remains a challenge – both getting it to form, and getting it to tip over.

Previous theories used a single giant impact to explain the tilt.  Uranus, though, has lower orbital eccentricity than the other gas giants save Neptune, and lower orbital inclination than the other gas giants.  It would seem strange such a collision would not have greater effect on the planet’s orbit or moons, which have relatively circular orbits in the orbital plane of Uranus.  Science Daily noted that problem for the single-impact theory, saying, “the moons of Uranus should have been left orbiting in their original angles, but they too lie at almost exactly 98 degrees.”  What to do?

This long-standing mystery has been solved by an international team of scientists led by Alessandro Morbidelli (Observatoire de la Cote d'Azur in Nice, France), who is presenting his group's research at the EPSC-DPS Joint Meeting in Nantes, France.

Morbidelli and his team used simulations to reproduce various impact scenarios in order to ascertain the most likely cause of Uranus’ tilt. They discovered that if Uranus had been hit when still surrounded by a protoplanetary disk – the material from which the moons would form – then the disk would have reformed into a fat doughnut shape around the new, highly-tilted equatorial plane. Collisions within the disk would have flattened the doughnut, which would then go onto [sic] form the moons in the positions we see today.

Problem: the model makes the moons end up in retrograde orbits, requiring additional tweaks to the model.  Adding two more impacts yielded a “higher probability of seeing the moons orbit in the direction we observe.”  Problem 2: “This research is at odds with current theories of how planets form, which may now need adjusting.”  If planets formed by accreting small material, the way textbooks have claimed for decades, then “They should have suffered no giant collisions,” Morbidelli said like a worried umpire.  Calling three strikes on Uranus will end the game: “So, the standard theory has to be revised,” he said – as if the planet-maker team was on base already (see 08/06/2004). 

In National Geographic's coverage, Morbidelli said that “So far, this is the only model that explains the equatorial orbits of Uranus's satellites.”  It comes at a cost of reconsidering the frequency of giant impacts: “But now we show that Uranus has to have been tilted at least twice, so these giant impacts were not exceptional events—they were the norm” – the norm only within the model, that is.  To accept the new model, other modelers will have to worry about creating more giant impactors and getting them to target one another.  Live Science indicated this does violence to the solar system base-running game: Morbidelli’s “monstrous collision” followed by two smaller collisions will have an impact on how planetary scientists must view the sun’s early innings: “The early solar system thus may have been a more volatile and violent place than previously thought, they added.”  That being the case, roller derby may be a more appropriate metaphor than baseball.

If you can tweak a contrived, oversimplified model and reproduce the real world, have you proven that’s what happened?  Has the real world been “solved,” as Science Daily claimed?  This is an important problem in philosophy of science – the use of models in theory construction. 

Modeling has a long and often successful history in science.  The Periodic Table was a model before it acquired enough empirical data to clinch it.  But that kind of modeling can be tested against real-time observation; modeling planets cannot.  The origin of Uranus, whatever happened, was a one-time, unique event.  For unique events, a model can only show what might have happened; it cannot show what did happen.  The plausibility of a model becomes suspect with (1) implausible initial conditions, (2) an increasing number of ad hoc parameters, (3) simplifying assumptions, and (4) bad consequences for other theories.  Think baseball.  Suppose you can model striking out a batter by imagining air guns along the way to push the pall into position.  That not only calls for contrived conditions; it changes the rules of the game.

Morbidelli (the guy who used a miracle to explain planetesimals; see 8/21/2009) wants us to believe his team has solved Uranus, but his model appears to suffer from all four weaknesses.  If Science Daily's coverage is accurate, he even claimed his model is a fact: "The fact that Uranus was hit at least twice suggests that significant impacts were typical in the formation of giant planets," he said.  Then he went on to require all the other planetary scientists to change their models because his model needs collisions: he needs three strikes, but their pitchers don't have the balls, "So, the standard theory has to be revised."  How's that for chutzpah.  But the multiple collisions necessary to create the Uranus we see appear so finely-tuned in his model, they appear almost miraculous (again).  If so, it’s doubtful that anything has been gained scientifically by the exercise except that Morbidelli can add another published paper, miracles and all, to his curriculum vitae.