Recapping the 2nd Workshop on Extreme Precision Radial Velocity
The 2nd Workshop on Extreme Precision Radial Velocity was held at Yale University July 5 - 8 with more than 150 attendees, including industry partners. The plenary sessions for this meeting were recorded and are available online here.
The first day of the meeting focused on the critical path challenges for reaching 10 cm/s precision from high resolution spectrometers. Significant progress has been made in the past few years on scrambling with optical fibers and broadband wavelength calibration with optical frequency combs. A single measurement precision of about 3 cm/s was demonstrated at the vacuum-enclosed and temperature-stabilized HARPS spectrograph using a Menlo Systems laser frequency comb (LFC) in April 2015, and there is every reason to believe that this will be the new standard in the next decade.
The meeting also featured "state of the art" presentations by the leaders of a dozen Doppler survey projects. These presentations showed that Doppler measurement precision improved from 3 - 10 m/s in the 1990's to a solid 1 m/s state of the art, circa 2015. A "hare and hound" fitting challenge was carried out before the meeting and revealed that synthetic signals with amplitudes smaller than 1 m/s were beyond our current reach - these signals were missed about 90% of the time. Radial velocity signals larger than 1 m/s were generally recovered, however the smallest of these signals currently require hundreds of velocity measurements to beat down errors with current state of the art measurement precision.
The next day of the workshop focused on statistical analysis techniques and a convincing case was made that the techniques employed by our community would not stand up to careful scrutiny of statistical reproducibility. The field of exoplanets has become sufficiently complex that no one person can be expected to have the expertise to solve all of the critical path challenges. The community was urged to fit for everything simultaneously, to tackle the new "fancy math" techniques, and to partner with experts in statistical analysis.
The long-standing question in the precision RV community has centered on whether or not it will be possible to distinguish stellar photospheric velocities from center of mass velocities. This was the topic of discussion on the third day and here the news is encouraging. There were demonstrations of substantial advancements, using principal component analysis of the cross-correlation function and other techniques.
There seemed to be consensus that we had not yet exhausted all of the tools at our disposal and acknowledgment that in order to detect Earth analogs, we must improve radial velocity precision and accuracy. The dream machine for Doppler planet searches will have very high resolution and very broad wavelength coverage to more efficiently identify stellar "noise." Putting this instrument on a dedicated, large aperture telescope will usher in a golden era for Doppler searches and significantly improve the efficiency and productivity of space-based missions.