Upon reading the headlines yesterday about the discovery of a liquid hot magma ocean within Io, my first thought was, "I sure hope Jason Perry writes something about this." I knew he couldn't stay away, and indeed he attempted to revive his long-dormant Gish Bar Times blog to post a lengthy writeup but was foiled by problems with Google's Blogger service. I'm all too happy to provide Jason with a bully pulpit for this exciting news from Io! --ESL
A fresh report was published online yesterday in Science Express on the discovery of a magma ocean beneath the surface of Io. I know! Big news! This is a paper I've been looking forward to seeing for more than year and half. I reported on this back in October 2009 when it was first presented at a Division of Planetary Sciences meeting and again in January 2010 when Richard Kerr wrote about that presentation, given by Krishan Khurana, in Science magazine. These results are now online as an article in press at the Science Express website (meaning the paper has been approved for publication but has not been published in the print version of Science). For those outside the Science mag paywall, check out the press release from the Jet Propulsion Laboratory.
Khurana and his co-authors, Xianzhe Jia, Margaret Kivelson, Francis Nimmo, Gerald Schubert, and Christopher Russell, re-examined Galileo Magnetometer data acquired during two of the spacecraft's encounters with Io in October 1999 and February 2000. Data was acquired on two other encounters, however they were polar passes and weren't nearly as useful for detecting a magma ocean. The magnetometer measured the absolute magnitude of the magnetic field surrounding the spacecraft and its magnitude in the three spatial components (Bx, By, and Bz). Near Io, the spacecraft mostly measured how Jupiter's magnetic field was perturbed by Io's atmosphere. In Io's atmosphere, plasma from Jupiter's magnetosphere slows as it takes on more mass and as charge is exchanged with these new particles. The magnetic field lines are also affected by interactions with Io's conducting ionosphere. Alfvén wings, which couple the Ionian and Jovian ionosphere in an electrical current called the Io flux tube, further affect the local magnetic field vectors (this charged particle interaction produces the aurorae seen at Io as well as the auroral footprint in Jupiter's atmosphere).
The big breakthrough came with the development of new magnetohydrodynamic (MHD) models in the last decade. These models allowed Khurana and his colleagues to sort out and remove all these interactions between Jupiter's magnetic field and Io's atmosphere and ionosphere. With these interactions removed, Khurana and his colleagues were able to look at Io's residual magnetic field, which they found had a field strength of greater than 500 nanoteslas. This residual magnetic field could either be induced by Jupiter's magnetic field, in a conductive layer within Io (like the field at Europa), or intrinsic to Io, generated by convection with Io's molten iron core (like the magnetic fields of Earth and Mercury).
So what could create an induced magnetic field at Io? Induced magnetic fields are created when a time-variable magnetic field sweeps through an electrically-conductive material, like the briny water oceans of Europa, Ganymede, and Callisto. Jupiter's magnetic field is tilted with respect to Io's orbital plane, so at times Io is above or below the normal plane of Jupiter's magnetic field. The time-variable magnetic field produces electrical currents within the conductive material, which produce a magnetic field through induction. The direction of this current changes twice each Jovian day (the magnetosphere rotates along with Jupiter itself, even at the distance of Io), causing the poles of the induced field to switch twice each Jovian day. Additional induced responses are created using the higher-order rotational harmonics of Jupiter's internal dynamo. [The point of this sentence is that Jupiter's magnetic field has a complex shape -- it's not just a simple bar magnet field -- and the lumps and bumps in Jupiter's magnetic field should have measurable effects in a field that they induce in Io. --ESL] (Scroll to the bottom of this post to see an animation of the interaction between Jupiter's field and Io.)
In order to determine the best fit to the available Galileo data, Khurana and his group created a model of Io's interior using multiple shells (like layers of an onion). They assigned each shell a conductivity based on its expected composition, temperature, and physical state, to measure the induced response to Jupiter's magnetic field. The interior model for Io used a bulk chondritic composition divided into a solid, cold, rocky crust 50 kilometers thick with zero conductivity and a molten iron core 900-1000 kilometers in radius. In between, they modeled Io's mantle with a composition similar to lherzolite, an ultramafic igneous rock found in Spitzbergen, Sweden and in the French Pyrenees; it consists of 44% SiO2, 32% MgO, and 14% FeO. The research team used experimentally-derived conductivity of lherzolite at various temperatures to simulate Io's mantle.
Khurana's group found that using a solid mantle, even one with induction, didn't provide a good fit to the Galileo data. They then added a conducting asthenosphere between the cold crust and the solid mantle. [The "asthenosphere" is a layer of rock within a terrestrial world that is able to flow or convect, as opposed to the "lithosphere," which is rigid, breaking rather than flowing. On Earth the lithosphere includes the crust and uppermost mantle; the asthenosphere is below that. --ESL] With the conducting asthenosphere (having a conductivity of 1 Siemen per meter or S/m) they were able to simulate an induced field with a strength greater than 600 nanoteslas, closely matching the Galileo data.
Is this a reasonable number? It turns out that like the saltwater in the subsurface oceans within Europa, Ganymede, Callisto, and Titan, ultramafic rock melts are also conductive with conductivities in the range of 1 to 5 S/m at 1200-1400°C. Partially molten rocks can have conductivities ranging from 0.0001 to 5 S/m. This approaches the conductivity of sea water, like the ocean found beneath Europa's crust.
How thick is this molten or partially molten conductive layer? It cannot be independently determined from the data, except to say that it is thicker than 50 kilometers. They did find that the induced field strength depends on how much of the asthenosphere is molten, and determined that Io's magma ocean would need to be at least 20% molten to replicate the Galileo data. So think of it as more a slurry rather than the ocean you might envision beneath Europa's surface, which would be much less viscous.
Finally, the authors determined that in order to produce this induced magnetic field, the magma ocean would have to be global, rather than just a few patches near active volcanoes or just along the equator, though they don't rule out variations in asthenospheric thickness or melt fraction due to differences in tidal heating between the equator and the poles.
This discovery does help to put to rest the question of whether Io has a magma ocean beneath its surface. You might assume it had one because of the widespread nature of its extreme volcanism. The idea was first proposed by M. H. Ross and Gerald Schubert in 1985 and it was revived in an Icarus note in 1999 by Laszlo Keszthelyi, Alfred McEwen, and G. Jeffrey Taylor. Since most of you probably don't subscribe to Icarus, you can read instead an article that Taylor wrote for the University of Hawaii website on this model.
Keszthelyi et al. proposed that the then-recent results from Galileo's Solid State Imager, suggesting that the lava erupting from Pillan in 1997 reached extreme temperatures of 2000 Kelvins, required a large melt fraction within at least the upper portion of Io's mantle. Their article presented a model where the melt fraction was approaching 35% near the boundary between the crust and the asthenosphere, and decreased the deeper you got into Io, until you hit 6% near the core-mantle boundary. However, the high temperatures seen at volcanoes like Pillan would suggest melt fractions in some places as high as 70%. A re-evaluation of the Pillan data in 2007 by Keszthelyi et al. reduced the eruption temperatures required at Pillan and conversely the melt fraction needed in Io's upper mantle to 20-30% with interconnected magma reaching down as far as 600 kilometers below Io's surface. The new results presented by Khurana confirm the presence of a magma ocean suggested by these authors and support Keszthelyi's current model of Io's interior.
Back to the latest results: after successfully modeling this induced field in Io's asthenosphere, Khurana used that model plus the magnetohydrodynamic model from the Galileo data to look for evidence for an intrinsic magnetic field at Io, one that would be created by convection within Io's molten core. They determined an upper boundary of 110 nanoteslas, making it a very weak magnetic field, if it exists.
What's next? To better determine the melt fraction within Io's magma ocean and to determine its thickness, more data is needed in order to resolve higher-order rotational harmonics from Jupiter's magnetosphere. However, even just two flybys worth of data has been enough to provide a useful proof of concept for probing the interiors of bodies like Io using electromagnetic sounding. By timing future encounters with Io to coincide with times where Io is not within the densest part of the Jupiter's plasma sheet, researchers would have an easier time picking out induced fields produced by weaker harmonics in Jupiter's magnetic field which may be lost in the noise of the moon/plasma interaction. If only there was a new spacecraft on its way to Io... another time perhaps.
Regardless, this is exciting news that Io's magma ocean has been independently confirmed by both eruption temperature data and models of Io's interior and by magnetic induction sounding. It is always nice to see Io get some press after all these years since the New Horizons flyby in 2007.