Jupiter's great red spot seen up close. Image credit: NASA/JPL-Caltech/ Space Science Institute

Jul 31, 2015 Solar Twin Planet Search finds a Jupiter analog

Sen—The Solar Twin Planet Search is a multi-national astronomy initiative to detect planets around stars similar to our Sun using the radial velocity (RV) method. They detect planets using the the High Accuracy Radial velocity Planet Searcher (HARPS) spectrograph on the European Southern Observatory (ESO) telescope at La Silla Observatory, Chile. They also use the Magellan Clay Telescope at Las Campanas Observatory, Chile to narrow their target list down to only Solar-twin stars—stars with temperatures, surface gravities (masses and radii), and elemental abundances extremely similar to our Sun. The overarching goal of the Solar Twin Planet Search (STPS) is to find a solar system analogous to our own.

STPS began their search for Solar twins with planets in 2009, and after narrowing down their target list to 63 Solar twin stars, they have detected their first planet around a Solar twin star. This planet just so happens to also be a near-twin to our own Jupiter. The planet, HIP 11519b, has a mass that is 99 per cent of Jupiter's mass and orbits its star at 88 per cent of Jupiter's orbital period. They note that based on the best fitting RV model, the possibility for terrestrial planets in smaller orbits is not ruled out for this system, which means the HIP 11519 system may be even more similar to the Solar System than we realize.

What makes the star HIP 11519 a Solar twin? STPS defines a Solar twin as having a temperature within 100 Kelvin of the Sun's temperature, a surface gravity within 25 per cent of the Sun's gravity, and a distribution of elements within 25 percent of the Sun's. HIP 11519 is well within those guidelines for a Solar twin: a temperature 10 Kelvin cooler, a surface gravity seven per cent heavier, and 13 per cent fewer heavy elements. This means that HIP 11519 is slightly smaller and about the same age as the Sun. It is also a closer Solar twin than Kepler-452, which was announced last week to have a super-Earth in the (optimistic) Habitable Zone. The moniker "HIP" means that the star was part of the Hipparcos mission to measure stellar distances. Based on the Hipparcos measurements, HIP 11519 is 2,838 light-years away.

The detection of HIP 11516b was made using the radial velocity method, which is considered by some to be the flagship method of exoplanet detection. The first exoplanet around a Sun-like star, 51 Pegasi b, was detected in 1995 by Michel Mayor and Didier Queloz using the RV method. 51 Peg was quickly followed by a dozen other exoplanets discovered the same way. It was not until the launch of the Kepler space telescope in 2009 that the transit method really took off, and other detection methods (including gravitational microlensing, pulsar and transit timing variations, and direct imaging) have not put out nearly the same number of discoveries.

The RV method detects an exoplanet by measuring the gravitational pull that a planet induces on its host star. A planet and a star mutually orbit each other around a common center-of-mass point, which means that while the planet is making its large orbit the star is making a much smaller orbit around itself. This is sometimes called the "stellar wobble," since the star is only moving a very tiny amount back and forth.

The stellar wobble is measured through changes in the stellar spectrum—starlight broken up into the contributions at individual wavelengths. Lines caused by the absorption of light at a particular wavelength will shift towards the red side of the spectrum if the star is moving away from the telescope (redshifting) and will shift towards the blue side if the star is moving towards the telescope (blueshifting). Since we can only measure the motion towards or away from us, not side-to-side or up-and-down, the measured motion is all in the "radial" direction, hence the name. An RV curve shows the repetitive redshifting and blueshifting of the stellar spectrum that indicates something is causing the star to wobble. The frequency, strength, and shape of the RV curve tells us the orbital period, mass, and eccentricity of the object's orbit around the star. Depending on these orbital characteristics, the object causing the wobble might be a planet.

The top shows a star and its planet orbiting around each other (the star's "wobble" is exaggerated for clarity). The bottom shows the redshifting and blueshifting of the stellar spectral lines from which we measure the radial velocity of the star and infer the mass of the planet. Animation credit: ESO/L. Calçada

Finding a Jupiter twin planet around a solar twin star with the RV method is not an easy task, for three main reasons. Firstly, a planet with Jupiter's orbital period will take over a decade to complete one orbit. That means that in order to fully trace the radial velocity curve we would need to measure the radial velocity of that star for over ten years at a good enough precision to detect a Jupiter-sized signal. Ideally those measurements would be spaced closely enough to see the RV of the star changing from redshifted to blueshifted. Even more ideal would be to see the full RV curve for more than one cycle, which would mean twenty years or more of RV data.

Secondly, there is the problem of the star itself confusing the RV data over such a long time period. Stars, including the Sun, go through stellar activity cycles which are the result of changing stellar magnetic fields. The Solar activity cycle causes events like Sunspots and large flares. These changes in the star's appearance can cause an RV signal themselves—if bright material from the star is being transported in the direction of your telescope this can cause a blueshift. The Sun's activity cycle occurs over an 11 year period—nearly the same as the orbital period of Jupiter. We would expect stars similar to the Sun to exhibit similar cycles, which complicates both measurements and analysis.

Finally, while Jupiter is the largest planet in the Solar System by far, it is still relatively small compared to the Sun, and is pretty far away. That means that the RV amplitude of Jupiter tugging on the Sun will only be about 10 meters per second. This is far from impossible for our current instruments, and we have detected planets that have much smaller RV signals, but again it is a matter of both time and precision. It is challenging to find enough data at this precision going far enough back in time to confidently detect a Jupiter-sized planet at a Jupiter-sized orbit. To date, we know of only three planets (including HIP 11915b) on orbits longer than 10 years with signals at 10 meters per second or smaller. Most planets detected with 10 year orbital periods induce much larger wobbles in their stars.

The STPS team was able to minimize each of these three problems, which allowed them to detect the RV signal from HIP 11519b. While STPS has only been in place since 2009, they were able to supplement their measurements of this star with archival data taken with HARPS going all the way back to 2003. This gave them a total span of 12 years of data, enough to fully detect a planet with the orbital period of Jupiter. Additionally, because the HARPS instrument is capable of measuring an RV signal with a nearly one meter per second precision, a planet with the RV strength of Jupiter can be measured accurately.

Determining the RV contribution from stellar activity was a much harder task. They based the amount of stellar activity on the spectral signature of calcium, which has some of the strongest spectral lines detectable. The shape, width, and depth of the calcium lines are largely dependent on variations in stellar surface temperature (starspots) and magnetic activity in the outer chromosphere (stellar flares). They measured the variation of the calcium lines simultaneously with their RV data and found no correlation between variations in stellar activity and the RV signal. Therefore it is really unlikely that stellar activity caused the RV signal attributed to the planet, but they cannot rule it out completely.

Why is it important that we find a Jupiter twin planet around a Solar twin star? If we really want to find a solar system analogous to our own, there needs to be a Jupiter-sized planet at a Jupiter-sized orbit. When the Solar System was forming, Jupiter came together first and essentially dictated how much of what material each of the other planets could gobble up. After the planets formed, Jupiter's mass and position determined how the orbits of the other planets interacted and evolved to their current positions. Without Jupiter exactly where it is, our Solar System would likely look radically different. If we expect to find a twin to the Solar System as a whole, it first needs to have a Solar twin star with a Jupiter twin planet. And now that STPS has shown that Solar System analogs are possible, we may find a solar system with all the matching components: star, gas giant planets, and habitable rocky worlds.


The RV curve of HIP 11519 using HARPS data over the course of 12 years (blue points) with the best fitting planet model (black curve) that indicates a Jupiter sized RV signal.