Last post I described the OHP2015 meeting in France, where I gave a talk.
My contribution to the conference proceedings just went live, and you can read them on the arXiv or on the OHP2015 site. For a “personal history” of the same material, told with greater authority and detail see Paul Butler’s post on the Pale Red Dot site.
The Lick Planet Search begin in the late ’80’s at the Shane 120-inch telescope on Mount Hamilton, in the mountains above San Jose (you can see the dome of the Great Refractor from the 101 freeway in San Jose).
The Shane 120-inch at Lick Observatory on Mount Hamilton, above San Jose. I spent many, many (too many) nights on that mountain in my many (too many) years in graduate school.
The spectrograph lives in a basement room beneath the siderostat shed seen here. On nights when other instruments were in use (like those using the laser, seen here), the spectrograph could still be accessed with an auxiliary telescope. The shed opens, and the siderostat reflects starlight up into the port in the side of the dome. A small, 0.6 meter telescope is housed inside the dome, and feeds the spectrograph by sending light down into the slit room in the basement, where a pickoff mirror sends the light into the path normally used for light from the 3-meter.
View of the siderostat from the perspective of the port in the side of the dome. This mirror feeds the small “CAT”: Coudé Auxiliary Telescope.
The work to extract precise RVs from the spectrograph was largely done at San Francisco State University by Geoff Marcy, Paul Butler, and their students and colleagues. The primary problem is that the instrument was not stable, and the “instrumental profile” or “line spread function” varied with temperature, pressure, position of its various components, and (trickiest of all) the illumination of the slit (which varies from exposure to exposure). Working with Jeff Valenti at CU Boulder, Marcy & Butler determined that they could use the precisely known spectrum of the iodine lines (obtained at the McMath FTS in 1991) to determine the instrumental profile simultaneously with the velocity of the star by modeling the spectrum. This was a significant computational burden at the time, but it worked.
Paul Butler at SFSU circa 1988
The Doppler lab at SFSU. The wall is lined with diagnostic plots from the Lick Planet Search.
The original cell was designed by Paul Butler and Geoff Marcy, and blown by SFSU glass blower Mylan Healy:
The original iodine cell for Lick.
Paul Butler, a chemistry masters student at SFSU, tested many different compounds to replace hydrogen fluoride, the original gas used by Bruce Campbell and Gordon Walker to get below 20 m/s precision at DAO. HF is notoriously dangerous to work with, and molecular iodine had been suggested by Robert Howard at the Carnegie Institution of Washington, inspired by work by Beckers in the late ’70s. Similar work was also being done by Libberecht and Hatzes & Cochran. Paul spent months trying many compounds many of them “explosive, deadly poisonous, or both” before settling on iodine.
Since pure iodine is a solid at STP, the gas cell is evacuated and heated to ensure complete sublimation of the small amount of iodine within. The blue cylinder in the picture is thermal insulation, and the wires are for the temperature controller that kept the cell at constant temperature. The arm held the cell in the path of the light.
Fancy picture of the iodine cell in position at Lick Observatory. I think that’s Debra Fischer’s hand. The beam is shown bouncing off of the slit plate, behind which is the spectrograph room. Image copyright Laurie Hatch, used with permission.
Another technical hurdle was a spectrograph that delivered high resolution, a broad bandpass, and a high-quality, linear detector. The Hamilton spectrograph fit the bill. It was built (and later upgraded) by Steve Vogt (one of Geoff Marcy’s thesis advisers) seen here holding a spooky glowing orb of some kind 1.
Paul calls the Hamilton “arguably the first modern echelle spectrometer”.
The Lick Planet Search really took off within days of the announcement by Mayor & Queloz of the discovery of 51 Peg b. Geoff and Paul immediately confirmed the discovery, devoting what little computing power they had access to to analyzing data just for that star from a single observing run. The ensuing attention gave them access to more computing power, and in short order they announced the next nine exoplanet discoveries (most already in their unanalyzed data!).
The Lick Planet Search spanned 25 years, and owes its success to almost innumerable observers (we tried to track down all their names and thank them in the acknowledgements here). Of special note are Debra Fischer, who ran and improved the program for years, and John Johnson, who did his thesis using the Hamilton.
The Lick Planet Search was still producing until one fateful day that the temperature controller partially failed. Unfortunately, the part that failed was the part that turned the heater off, and as a result the insulation was melted and the cell badly damaged.
The sad fate of the Lick iodine cell.
My understanding is that Geoff Marcy donated the cell to the Smithsonian.
In total, the Lick Planet Search made 14,000 observations of 386 stars. Debra published the entire radial velocity archive in 2014, ensuring that its legacy will be maintained.
Next post: the planet search moves to Keck.
1 Yes, I know it’s a mirror.