A Hawaiian interlude: part 2

Request OA slew to new field, move grating wheel to mirror position, move out slit-mask, put in filter, change instrument focus, set exposure time to 20s, take exposure, read out detector, verify new pointing against finder chart, identify set-up star, get its coordinates from image, move in slit-mask, take 20s exposure, read out detector, identify set-up box in image, command coarse telescope offset to align slit-mask, take 20s exposure, fully read out detector, initiate alignment script, identify star alignment boxes, command script to measure star alignment with reference to set-up boxes, compute fine telescope offset and rotation, request OA to stop guiding, apply offset and rotation, request OA to resume guiding, take 20s exposure, read out detector, run alignment script, repeat whole fine alignment procedure to tweak position, or (if last fine offset and rotation were good) move in grating, move filter wheel to clear, change instrument focus, set grating angle, set exposure time to 1800s, begin integration. Phew....

The observational astronomer in his native habitat.

Well, it's typically an 18 to 20 hour day. You arrive soon after lunchtime and talk to your support astronomer - a local staff scientist - to find out about any current problems and quirks of the telescope and instrument that you will be using. There usually are several. Next you go over the plan for the observing run or - if its your 2nd or 3rd night of observing - revise plan based on what you had accomplished the night before. If you still have time, you may take a closer look at the data you took last night to make sure all's well. Around 4 PM the day crew working at the summit release the telescope to you and the fun begins when you start calibrating your instrument. This usually takes a couple of hours. If you are lucky, you may squeeze in dinner. I usually bring a sandwich.

Greg Wirth - an old acquaintance - here seen in a rare moment of rest. Greg is a staff astronomer at Keck and was my support astronomer for this observing run. His job is essentially to make sure that all my observing needs are taken care of and to fix (or find out who can) all the problems that develop with the instrument I am using. Great service with a smile.

Dave, one of the Keck techs, up at the summit. It's guys like him get the telescope and instrument ready for the night, each night of the year. It takes about 100 people to run a modern observatory like Keck.

In the olden days all the observing was done from Mauna Kea summit. Nowadays, the astronomer gets to observe remotely from Waimea where there is a lot more oxygen for the brain. Only the Observing Assistant, who is responsible for keeping the telescope healthy, is at the summit.

Closer view of the instrument control computer down at the remote observing facility in Waimea. Through this interface the astronomer controls the instrument and how the data are taken

Carolyn, the OA (Observing Assistant), keeping an eye on things up in the Keck I control room at the summit. We have a continuous video link going throughout the night. The OA at the summit controls the telescope while the astronomer down at the remote observing facility in Waimea concentrates on taking the data

Officially, the night begins at sunset when the OA opens the dome shutter and the telescope first sees the sky. For the next half hour or so it is still way too bright to do science, but a number of important procedures get done, including focusing the telescope and making sure all is functioning as it should be. When it gets dark enough, you can start observing your science targets and the fun begins. It's usually a busy night, especially when things start going wrong with the instrument, when the weather start going funny, or when you start making mistakes. More often than not at least one of these things happens. Meanwhile, time is critical, with the clock ticking and there being only so many minutes before the sun rises again. It costs about $1 a second to operate a modern world-class telescope, so you don't want to waste too much of that time by being silly or inefficient. And if you run out of time and don't take the data you need, you will have to compete again to get more telescope time to finish your project. You certainly don't want that and so you better stay sharp and focused, sacrifice a graduate student to the weather gods, and in general hope it all goes well. For the most part, on this observing run, it did. There were some problems with the instrument, and one or two minor mess-ups, but all in all, I got really good data.

The nature of the project I was working on (deep spectroscopy of faint sources) was such that the night consisted of relatively short moments of stress and activity, followed by intervals of relative peace. During the stressful moments - when I was setting up the next set of targets - lots of things needed to be done in just the right order but with lots of critical decision points along the way (see first paragraph of this post). After the target was set up, relative peace would ensue while a long, 30-minute integration was underway. But that peace did not necessarily mean lack of activity: data from the last integration needed to be checked and verified, and decisions often had to be made about what to do next with the shrinking time remaining in the night.

Overall, this observing run was pretty exhausting: often you share the work with another astronomer, a colleague who's working with you on the project. But this time I was alone. So the nights were long and busy. Fortunately, during this run I had some friends nearby who would pop in to keep the morale up during the less busy moments...

Observing over on Keck II were two comrades from my Caltech days. Keith Matthews (left) and Dave Thompson (right)

And it sure was nice that I got some good data!

That, my friends, is the spectroscopic signature of a Lyman-alpha emitter at redshift z=5.7. Doesn't look like much, you say? But consider that it's most likely a baby star-forming galaxy 12.65 billion light years away. [And in case you were wondering, that'd take you 19888186 years to cover even at warp 8.6.] So it's pretty darn far. This galaxy - one of several I got on this observing run - emitted its light at a time when the Universe was less then a tenth its present age.


As the night ends, it's almost time to go to bed... but not yet. After the telescope dome closes, it's time to do some final calibrations that are tedious but vital for making the data scientifically useful. These may often take another hour or more, sometimes stretching "the night" well into daylight hours. With the night over you go to bed around 7 or 8 am only to get up around lunchtime. But if you are lucky, and it's the end of your observing run, you get to sleep in and then go to the beach before flying home the next day...

I got to go to the beach the evening after my run ended. And then it's back to Kyoto the next day.

A Hawaiian interlude: part 2

Request OA slew to new field, move grating wheel to mirror position, move out slit-mask, put in filter, change instrument focus, set exposure time to 20s, take exposure, read out detector, verify new pointing against finder chart, identify set-up star, get its coordinates from image, move in slit-mask, take 20s exposure, read out detector, identify set-up box in image, command coarse telescope offset to align slit-mask, take 20s exposure, fully read out detector, initiate alignment script, identify star alignment boxes, command script to measure star alignment with reference to set-up boxes, compute fine telescope offset and rotation, request OA to stop guiding, apply offset and rotation, request OA to resume guiding, take 20s exposure, read out detector, run alignment script, repeat whole fine alignment procedure to tweak position, or (if last fine offset and rotation were good) move in grating, move filter wheel to clear, change instrument focus, set grating angle, set exposure time to 1800s, begin integration. Phew....

The observational astronomer in his native habitat.

Well, it's typically an 18 to 20 hour day. You arrive soon after lunchtime and talk to your support astronomer - a local staff scientist - to find out about any current problems and quirks of the telescope and instrument that you will be using. There usually are several. Next you go over the plan for the observing run or - if its your 2nd or 3rd night of observing - revise plan based on what you had accomplished the night before. If you still have time, you may take a closer look at the data you took last night to make sure all's well. Around 4 PM the day crew working at the summit release the telescope to you and the fun begins when you start calibrating your instrument. This usually takes a couple of hours. If you are lucky, you may squeeze in dinner. I usually bring a sandwich.

Greg Wirth - an old acquaintance - here seen in a rare moment of rest. Greg is a staff astronomer at Keck and was my support astronomer for this observing run. His job is essentially to make sure that all my observing needs are taken care of and to fix (or find out who can) all the problems that develop with the instrument I am using. Great service with a smile.

Dave, one of the Keck techs, up at the summit. It's guys like him get the telescope and instrument ready for the night, each night of the year. It takes about 100 people to run a modern observatory like Keck.

In the olden days all the observing was done from Mauna Kea summit. Nowadays, the astronomer gets to observe remotely from Waimea where there is a lot more oxygen for the brain. Only the Observing Assistant, who is responsible for keeping the telescope healthy, is at the summit.

Closer view of the instrument control computer down at the remote observing facility in Waimea. Through this interface the astronomer controls the instrument and how the data are taken

Carolyn, the OA (Observing Assistant), keeping an eye on things up in the Keck I control room at the summit. We have a continuous video link going throughout the night. The OA at the summit controls the telescope while the astronomer down at the remote observing facility in Waimea concentrates on taking the data

Officially, the night begins at sunset when the OA opens the dome shutter and the telescope first sees the sky. For the next half hour or so it is still way too bright to do science, but a number of important procedures get done, including focusing the telescope and making sure all is functioning as it should be. When it gets dark enough, you can start observing your science targets and the fun begins. It's usually a busy night, especially when things start going wrong with the instrument, when the weather start going funny, or when you start making mistakes. More often than not at least one of these things happens. Meanwhile, time is critical, with the clock ticking and there being only so many minutes before the sun rises again. It costs about $1 a second to operate a modern world-class telescope, so you don't want to waste too much of that time by being silly or inefficient. And if you run out of time and don't take the data you need, you will have to compete again to get more telescope time to finish your project. You certainly don't want that and so you better stay sharp and focused, sacrifice a graduate student to the weather gods, and in general hope it all goes well. For the most part, on this observing run, it did. There were some problems with the instrument, and one or two minor mess-ups, but all in all, I got really good data.

The nature of the project I was working on (deep spectroscopy of faint sources) was such that the night consisted of relatively short moments of stress and activity, followed by intervals of relative peace. During the stressful moments - when I was setting up the next set of targets - lots of things needed to be done in just the right order but with lots of critical decision points along the way (see first paragraph of this post). After the target was set up, relative peace would ensue while a long, 30-minute integration was underway. But that peace did not necessarily mean lack of activity: data from the last integration needed to be checked and verified, and decisions often had to be made about what to do next with the shrinking time remaining in the night.

Overall, this observing run was pretty exhausting: often you share the work with another astronomer, a colleague who's working with you on the project. But this time I was alone. So the nights were long and busy. Fortunately, during this run I had some friends nearby who would pop in to keep the morale up during the less busy moments...

Observing over on Keck II were two comrades from my Caltech days. Keith Matthews (left) and Dave Thompson (right)

And it sure was nice that I got some good data!

That, my friends, is the spectroscopic signature of a Lyman-alpha emitter at redshift z=5.7. Doesn't look like much, you say? But consider that it's most likely a baby star-forming galaxy 12.65 billion light years away. [And in case you were wondering, that'd take you 19888186 years to cover even at warp 8.6.] So it's pretty darn far. This galaxy - one of several I got on this observing run - emitted its light at a time when the Universe was less then a tenth its present age.


As the night ends, it's almost time to go to bed... but not yet. After the telescope dome closes, it's time to do some final calibrations that are tedious but vital for making the data scientifically useful. These may often take another hour or more, sometimes stretching "the night" well into daylight hours. With the night over you go to bed around 7 or 8 am only to get up around lunchtime. But if you are lucky, and it's the end of your observing run, you get to sleep in and then go to the beach before flying home the next day...

I got to go to the beach the evening after my run ended. And then it's back to Kyoto the next day.