A visit to Apache Point

Last week, I had the opportunity to visit Apache Point Observatory, where the photons providing data for the Sloan Digital Sky Survey are gathered safely to Earth and their final digital form. The observatory is situated at the edge of a ridge in the Sacramento Mountains of New Mexico, near the solar observatory complex at Sunspot. (Back when that was built, the astronomers obviously had a hand in picking the route number of the highway leading from the town of Cloudcroft – spectrum aficionados will recognize the New Mexico highway number). At its elevation of 2800 meters (9200 feet), there were still patches of snow despite being able to look out across desert below. The steep western face of the mountains looks out across the White Sands, both the gypsum-rich dune field and the historic rocket test range of the same name (and the landing strip once used by a space shuttle when the weather was terrible on both coasts of the US).

On arrival, site manager Bruce Gillespie gave a detailed and gracious tour of the site (freely pointing out some of the engineering and operational details of which they are justifiably proud, just the kind of thing I’m a sucker for at observatories). He was primed for the Zoo connection – just the previous week, the daily radio spot “StarDate” had an episode on Hanny’s Voorwerp, so he recalled the Zoo and “something about a Dutch schoolteacher”. Four telescopes share the site, lined up along the edge of the mountaintop. This location, where the prevailing wind comes whipping up the steep slope and across a small flat plateau, compresses the most turbulent layers of airflow so much that telescopes can be put on stilts to get above them (which explains the odd form of the Sloan telescope building). The first telescope erected at the site was the 3.5m telescope of the Astrophysical Research Consortium, one of the first of a wave of university groups building telescopes from the 1980s onward. This instrument may be best known for followup spectra of the highest-redshift quasars found by the Sloan survey, but it is remarkably flexible with physical instrument changes taking only a few minutes and being routine during a night. (It even fires laser pulses at the lunar retroreflectors). Many observers use the 3.5m telescope remotely from Seattle, Chicago, Princeton, Las Cruces… Also located here, in small domes raised on pillars, are the 1-meter telescope of New Mexico State University, and the SDSS photometric telescope. This 0.5m instrument monitors the brightness of standard stars all over the sky so the main telescope doesn’t need to interrupt its program to do so, cross-calibrating any small changes in atmospheric transmission during the survey scans.

The SDSS 2.5m telescope is controlled from the same building as the 3.5m. When the camera is in use, a wall-sized bank of monitors shows the images scrolling by as all 30 CCD chips scan the sky. The telescope itself is some distance away, with a giant roll-off shed leaving it exposed to the wind (and quickly matching the temperature of the surrounding air) during operation.

During our visit, the wind was blowing strongly, so the best view of the telescope came when Bruce was able to raise only the downwind door for a few minutes. This was a good chance to show the Galaxy Zoo flag…

The telescope’s signature feature is the corrugated wind baffle, the box surrounding the whole telescope. The baffling stops the wind from shaking the telescope, but lets the air flow gently through to keep temperatures equalized and minimize internal turbulence. The wind baffles are on a separate mount, unconnected to the telescope mount. The telscope, like so many new-technology instruments, uses an altazimuth mounting, with simple pivots in elevation and azimuth (compass bearing). This means that it works only under constant and smooth computer control of movements in both axes, but gives a much smaller, lighter, and cheaper design than an equatorial mount which counteracts the Earths rotation automatically in an analog way. Not only does the telescope have to move constantly under computer control, but the camera or spectroscopic apertures must do so as well, since the sky rotates in the telescope’s field of view while the Earth turns. The imaging surveys for SDSS are essentially finished. The last 2000 square degrees was recently completed, extending the SDSS footprint a bit farther south and filling in the patch of northern sky missing for Data Release 7. Use of the telescope is switching toward full-time spectroscopy, overlapping several different surveys (and adding some instruments). One (BOSS) will concentrate on the distribution of galaxies and quasars, driven by the so-called baryon oscillations (a relic of the early Universe which expands with it, giving us a new way to measure its expansion history). This project will target galaxies and quasars at a wide range of redshifts, and start using longer exposures (deeper galaxy spectra – mmmmmmm). The ongoing SEGUE project centers on stars in our galaxy, to be extended into SEGUE-2, and a new infrared spectrograph will map the Milky Way using stars detected in the infrared by 2MASS (APOGEE). Finally, a new high-dispersion spectrograph will be located in the basement of the adjacent building and be able to search for the Doppler motions caused by planets around over tens thousand of stars (MARVELS). The heart of the ability to measure hundreds of spectra at once is the use of optical fibers. These connect the positions of objects in the telescope to a pair of spectrographs (green boxes in the picture of the back of the telescope), allowing them to see light from stars and galaxies across the mammoth 3-degree field of view.
For each pointing of the telescope, there is an aluminum plate with precisely drilled holes to match the desired objects. These holes are much larger than the 3-arcsecond area of sky to be measured, to allow for the protective jacket and connector of each optical fiber. That size makes them very easy to see when lit up, for example by light coming through a window: This metal aperture plate is attached to one of nine fiber cartridges, each of which has its own set of optical fibers bundled at one end into the narrow slits for the spectrographs.
The loose ends are plugged into the holes in the plate (by people whose job title is “plugger”, of course). They don’t have to keep track of exactly which fiber goes where – since it would be ridiculously time-consuming for people to do the necessary double-checks, an automated system including a laser and TV camera spends five minutes working out which fiber goes where, and tracks the bookkeeping. Nine of these cartridges give enough to keep the telescope busy for the survey on a long winter night; they are all replugged during the day once a given field has been successfully observed. Over 2500 of these plates have been fabricated – after handing them out to interested parties and museums, nearly 4/5 were recycled (shouldn’t GZ have a memorial souvenir spectrum plate? I should have asked while there was space in the trunk…)
Just to be sure the survey staff remembers Galaxy Zoo, I not-so-surreptitiously left that banner on one of the office bulletin boards in the control room. Just a reminder of 182,000 satisfied SDSS data customers.

Next stop on this trip – Roswell, New Mexico. Here I was able to document a little-known fact about the aliens’ mission to Earth. If you look carefully, you can see what they came for!