paper: ps, pdf

Menu:

Travel 2009:

Jan 04-08: Long Beach, CA
Talk at AAS Meeting

June 19-30: Prato, Italy
Linfest

July 1-5: Tenerife, Spain
Talk at IAC

July 12-Aug 24: Princeton, NJ
PiTP 2009: Computational Astrophysics

July 30-Aug 8: La Serena, Chile
Observations at CTIO

Aug 10-15: Rio de Janeiro, Brazil
IAU General Assembly

 Terrestrial Planet Formation Around Alpha Cen B

Javiera Guedes, Eugenio Rivera, Erica Davis, Greg Laughlin, Elisa Quintana, and Debra Fischer.

  Alpha Centauri B is an excellent candidate for hosting Earth-sized planets. Its mass is 0.9 MSun and its metallicity is similar to that of HD69830, a quiet K0 star host of three Neptune-like planets. Alpha Centauri B is the binary companion of Alpha Centauri A, a G2 star of mass 1.1 MSun. Proxima Centauri, a red dwarf is believed to orbit the AB system with a semi-major axis of 13,000 AU.
We study the formation of Earth-size planets around Alpha Centauri B by evolving a disk of Moon-size oligarchs using John Chamber's code mercury.f, a hybrid symplectic code that accounts for the presence of the binary companion. Though most of the accretion occurs within the first 70 Myr, we allow the integration to run for 200 Myr to assure the stability of the final planetary configurations.

Run 07: N = 700 Run 08: N = 600

Click in the picture to run the movie.

Are these planets observable? The short answer is Yes! If they exist, we can observe them. The left panel below shows the synthetic radial velocity profile of the four planet system created in run 08. The masses of these planets are M = 0.07 MEarth at a = 0.2 AU, M = 0.6 MEarth at a = 0.7 AU, M = 1.8 MEarth at a = 1.09 AU, and M = 0.6 MEarth at a = 1.8 AU. The star's radial velocity has an amplitude of only 0.3 m/s.
We add a (white) noise level of 3 m/s (ten times larger than the signal) and take synthetic data every 200 seconds for 5 years (right panel). These data are based on the weather condition of the Las Campanas Observatory. The thin streaks in the data are purposely added in to represents days at which data cannot be taken (power outages, bad weather conditions, etc.) while the 60-day gaps account for the two months a year when Alpha Centauri B is too low in the sky (airmass > 3). Based on the fact that the d planet in the HD69830 system was found with ~75 observations, a 2 MEarth planet would need ~500 more observations. This amounts to a minimum of 35,000! The data generated below (right) has a far larger number of observations.

Note the difference in scale between the two panels above. The star's radial velocity signal is much weaker than the noisy data set we synthesized. Can the signal be recovered and the planets detected at this noise level? The panel below shows the evolution of the power spectrum as more and more data are taken. The inset shows the synthetic data taken as a function of days of observation. After 5 years, the power spectrum shows a clear peak at the period of the larger planet!

Click in the picture to run the movie.

But what it the noise was not white? According to the GOLF experiment on the SOHO satellite, the radial velocity of the Sun is quite noisy and seems to increase in amplitude over time. If this noise is not a detector artifact but it's due to solar activity, then it would be very hard for an extra-solar civilization to detect our Jupiters and Saturn, not to mention Earth. In order to study the effect of noise in the observation of Earth-size planets, Greg added noise at the GOLF-measured level to the radial velocity curve of HD69830. Systemic users fit this curve but could not find the three already detected Neptunes present there. This would mean that the sun is far noisier than HD69830 (and twin starts such as Alpha Centauri B), and therefore we don't expect that the noise from the star will stop us from finding its Earths. For more information, read Greg's Oklo Post.