KITP: Exoplanets RIsing

Well, I’m back at the Kavli Institute attending the Exoplanets Rising workshop.

We have a full schedule of talks over the week, and I’ll be intermittently blogging the events as we amble along.
Bunch of interesting sounding talk on the schedule, and hopefully some interesting news and discoveries that we will hear about.

We kick-off this morning with Mayor and Marcy, and the Fischer and Charbonneau after the break, looking forward to it.
Not seen planets around M dwarfs with mass less than 0.3 solar masses, but that is likely a bias – lower mass dwarfs are fainter, duh.

Talks will be online, as usual, both video, podcast and slides (eventually).

Previous talks from program are here.

Mayor: HARPS southern search looking at long period giant planets in metal deficient systems. Found 3 new planets around stars with -2 < Z < -0.5 out of a sample of 100.
Looks like a flat tail of long period (> 1 yr) giants around moderately metal poor stars (Z ~ -1 or so). Then the well know increase for metal rich stars (P ~ Z2 for short period giants and -0.5 < Z < +0.5)
- Been expecting that, was going around for years saying there ought to be more than predicted long period giants around the metal poor stars...

Search for stars around intermediate mass stars (1.5 - 5 solar mass planets).

Surveys to tackle larger samples of stars using multi-object spectrographs,
eg. survey tacked onto Sloan III survey.

HARPS southern survey of 400 brightest M dwarfs. Looking for Neptune mass planets and Super-Earths. Finding planets.
Search confounded by spots on slowly rotating stars.

HARPS now spending half its time on 250 "good stars" looking for low mass planets, going to very high velocity precision, high cadence and long timeline for observing - going since 2004. Getting to below 1 m.sec
Finding good stuff.

eg. HD181433
0.7 M_Jup at 1024 days
and 7.5 M_Earth at 9 days
going too fast for me to get the numbers but they are all published

Timescales disparity in multi-planet system, need high cadence, expensive.
Still no optimizing strategy for sampling - strange.
Seeing LOTS of low mass planets at all radii where there is sensitivity.

At least 30% and likely 50+% of stars have Neptunes/Super-Earths.
Many systems with multiple Super-Earths.
Found a 5 planet system.
Seeing many systems with significant eccentricities. - Scattering common? Systems form dynamically crowded.

Seeing lots of resonant orbits also - consistent with dynamical crowding.
- Makes both data analysis and observing strategy optimization harder.

Preliminary hints:
N(planet) ~ 1/M2 approx for M < 40 Earth masses. Maybe.
Low mass planet numbers at orbital period looks to be rising for P > 10-30 days.
No metallicity correlation for low mass planets ( M < 10 earths). Hmm.
Caveat: serious observing biases. Sample chosen for quietness, not uniformity, and dex spread only about 0.6

HARPS getting down to 0.1 m/sec on bright stars (ie α Cen).
Can already detect Earth’s in habitable zones of nearby bright stars with long timeline high cadence observations.

Targeting 10 bright nearby quiet main sequence stars in the south.


  1. #1 andy
    March 29, 2010

    There are already some talks online as part of the Theory and Observation of Exoplanets programme.

    Talking of low-metallicity environments and planets, are there any surveys aimed at finding long-period planets in globular clusters, or is that too challenging at present without convenient millisecond pulsar binaries?

  2. #2 Steinn Sigurdsson
    March 29, 2010

    thnx – put up link to program talks

    There was an infamous transit search of 47 Tuc little over a decade ago, a week with HST, didn’t find anything but started a lot of interesting things, including demonstration of what could be done with transits and from space respectively…

    There is an Australian group that has looked for short period transits in the outskirts of 47 Tuc, also with no luck.

    The long period planets are impossible in globulars for now – too crowded and individual stars are too faint.
    Unless you have a handy dandy millisecond pulsar as you say.
    Maybe doable with a multi-object spectrograph and 30m telescops, if you could convince a TAC to let you use it a lot for a long time and had a suitable hi-res spectrograph on it.