continued from
last week.
After lunch comes the final push for getting ready for the night. The support astronomers (astronomers who work at the observatory and who know much much much more about the telescopes and instruments than you could ever hope to know) arrive to help start setting things up for the night, the daytime crew at the summit does their final checkouts in preparation for the evening, and the astronomers (Emily and I) do the final mental and computational run through of the night to make sure everything is going to work out OK.
Emily and I are both paranoid about having done something dreadfully (or, even, trivially) wrong that will result in the loss of observing time. I think that such a healthy paranoia is one of the best traits an astronomer could have. Chances are, in fact, that you
have done something dreadfully wrong the first time, and only by being paranoid enough to double, triple, and (with the two of us trying to out-paranoid each other) quadruple check do you actually get anything done. Today our paranoia results in the discovery of a bug in a computer program that we use to track the Kuiper belt objects across the sky. The bug causes the telescope to track a small fraction slower than it should. We fix the bug, do all of the calculations on paper with a calculator to make sure we really are getting the right answer, and then convince ourselves that we are indeed ready to go for the night.
Meanwhile, Jim, one of our support astronomers, walks in the room and casually says "oh hey we may be a little slow firing everything up this afternoon, as there was a power glitch at the summit just a few minutes ago."
Usually power glitches don't do anything, as most of the observatory runs on backup for short glitches, but, as we just learned, the laser, in particular doesn't. If I understood the explanation correctly, the laser almost draws more electrical power than the entire rest of the observatory. It has to be isolated on its own circuit. Luckily, the Hawaii Electrical Company (HELCO. At some point in the afternoon someone remarked "Add one more L and the name would be appropriate) essentially guarantees that there will be no glitches. Unluckily, this guarantee does little to prevent glitches.
Apparently, glitches are bad for the laser. I would tell you why, but I don't actually understand enough to tell you why. I am not an engineer. I tried building things with my brother -- who is an engineer -- when we were kids, and his were always fabulously designed and assembled constructions, while mine would always hang precariously for a while until they fell to the ground. At which point I could explain, in theory, why my design should have been perfectly good. Astronomers rely on a vast pyramid of highly skilled and highly creative people who understand all of these things we don't (laser physics, oil bearings, active mirror control, cryogenic mechanical operations, to name a tiny fraction) and who are dedicated to making them work.
The laser was not coming back up.
In any complex system like a Keck telescope many many things have the possibility of going wrong, so it is not unusual for there to be a little afternoon panic. We've learned not to panic. Emily and I, now fully prepared for the evening, spent our time watching the satellite weather image update every half hour. The huge mass of clouds to the west was slowing blowing our way, pixel-by-pixel. Based on how the clouds had moved over the past twelve hours, we were pretty sure that we would have a few good hours in the evening before they hit. And, even then, you never know. We have been at the telescope before when an impenetrably opaque bank of clouds mysteriously parted just as it reached the Big Island. So we still had hope.
In the background we could hear conversations taking place at the summit, between the summit and the small crowd gathered in the control room, and between some in the crowd and other people who had been called. It was a Saturday. The laser was not coming back up.
The first line of defense when something goes wrong is the set of people who are working that day. Yes, it was a Saturday, but Saturday had a night like any other day, so there were people working hard to make sure the telescope worked that night. But when the first line of defense fails, anyone and everyone can be called at any time of the day or night. In the background conversational snippets I could hear, I started to hear a common refrain. "I think we need Kenny." "Has anyone seen Kenny?" "Do you know where Kenny is the weekend?" Voices were beginning to sound slightly more panicky.
Kenny knows the laser better than anyone. If the laser is acting up, Kenny can calm it down. If the laser is not coming back up, Kenny will know what to do. But Kenny was not on call on Saturday. If someone is on call they are required to be a phone call away. Kenny could be anywhere on the island or just asleep with his phone unplugged. There was no way to know.
It was about this moment when someone slipped into the back of the room. He was wearing beach apparel, slightly sunburnt, and, frankly looked like he hadn't showered in a few days. If it weren't for the fact that I know that the building is locked, I would have guessed that someone had hiked the 15 miles uphill from the beach and was looking for a cool glass of water. Or, more likely, beer. Everyone turned around: "Kenny!"
"Sorry for the clothes, I've been out camping and just got the message on my cell that the laser is not coming up."
Everybody seemed greatly relieved. The voices all around continued. Emily and I went over plans and counter plans. The mass of clouds continued its slow march.
About an hour and a half before sunset, the Observing Assistant arrives at the summit. The OA is the one who drives the telescope all night long and makes sure everything goes smoothly at the mountaintop. Another piece of the pyramid. And, tonight, it is Cindy, one of our favorite people to work with in the whole place. Emily and Cindy and I have spent many nights together at the telescope (or at least she is at the telescope; we're still down in Waimea), sometimes working hard, sometimes trying hard to stay awake, sometimes playing silly games to distract ourselves from the fact that the night is slipping away and we're still not getting any work done because it is raining on the telescope.
As sunset closed in the conversations surrounding Kenny sounded up beat. The laser was almost there. It was going to be yet another afternoon of panic followed by smooth sailing all night long.
"What was THAT?" said Kenny.
Not a good question for Kenny to ask.
He asked Cindy, up at the summit: "Was there a power glitch?"
"Well, yeah, this afternoon," Cindy replied.
"No, I mean right now," said Kenny.
Cindy checked all of her systems, called other telescopes, looked everywhere, and finally declared, no. No glitches. This time, for once, HELCO did not do anything it wasn't supposed to.
But the laser was dead. And no one knew quite why. And to make things worse, the sun set. On a good night, we would, right now, open up the telescope, swing the dome around, and begin the night's work. Tonight, we were dead in the water, and there was nothing Emily or I could do. We watched the satellite image, recalculated the time that we thought the clouds would hit, and listened to the chatter in the background. We were not panicking at this point, but we were suddenly feeling extremely antsy. The sky was still clear for maybe two more hours, and our telescope was not even open.
We assumed that the laser would come up pretty quickly when suddenly one of the support astronomers asked, "Are you sure there were no power glitches? The entire adaptive optics system has lost power, too."
This whole adaptive optics system is sufficiently complicated that every single day it needs to be carefully calibrated to make sure that it is correcting the distortions of the atmosphere just right. Losing power meant that all of those calibrations had to start from scratch. Our first support astronomer, Jim, now joined by a second, Hien, set to work doing the recalibration.
Phone calls were made. Saturday night dinners were ruined. Evenings out were postponed. Everyone who knew things about the laser and the adaptive optics were called in from whatever they were doing.
I can't relay all of the different strings of conversations that were going on throughout the room. Neither Emily nor I were one bit of use to this, so we just stayed in the background, talking quietly, watching the clouds get closer and closer. At some point Kenny realized that not only were his Saturday night plans slightly modified, they might soon take a colder turn. Nothing was working and no one knew why, so Kenny decided he was probably going to have to take the hour long drive up to the summit for some hands-on work. The summit temperature hovers around freezing, so Kenny decided it would be best to take a shower and find some warmed clothes before he went. No one is ever really prepared for having to go to 13,000 feet when just a few hours earlier they were camping on the beach.
Jim warned us: "This is going to take at least an hour, in the best possible case. And I mean an hour from whenever the power is restored." Which was not yet. "Would you like to switch to a different instrument?"
As the sky was clear, Jim was offering us the chance to do something
other than adaptive optics, which was better than doing nothing at all. We jumped up from our corner.
"Let's switch to NIRSPEC" I said to Emily.
NIRSPEC cleverly stands for near-infrared spectrograph, which means that we could use the instrument not to take extra-crisp pictures, but to analyze the light that comes from the object to see what the object might be made out of.
Switching to NIRSPEC was not in our contingency plan. The possibility that the sky might be clear but the adaptive optics system might be broken was so far down our list of possibilities that we never planned. Emily and I ran down the possibilities.
Me: "We can only look at the brightest of the Kuiper belt objects with NIRSPEC. What's up in the sky right now"
Emily: "2003 EL61 won't rise for about another hour. Quaoar will not rise until the end of the night. Orcus and 2005 FY9 are both up right now."
Me: "OK, we will not be able to do anything useful on Orcus in a short amount of time. I'd love to have NIRSPEC on Orcus for 3 nights in a row. But a few hours? Worthless."
Emily: "Agreed. OK. 2005 FY9? We have already analyzed much of the spectrum. We could collect more data to simply add it to what we have."
Me: "Boring. Let's try something new. There is a region of the spectrum that we've never looked at before. We never looked because we assumed there was nothing interesting there. But wouldn't you rather look somewhere new on the off chance that there might be something interesting that you hadn't thought about before than to look somewhere old to see it slightly better when you already know what is there?"
Emily: "That is only a moderately good argument."
Me: "Tonight I will definitely settle for moderately good."
Me, to Cindy: "Let's take the telescope over to 2005 FY9."
We got to the nice bright Kuiper belt object, began to collect data, and started to relax a little. The voices in the background were still a bit frantic. Even more people were in the room. But at least we were collecting new data.
Ten minutes after we started, the first cloud hit.
"Where did our object go?" Emily asked.
Cindy went outside to look, and it was true. Some of the clouds were arriving early. Even our desperation backup plan was not going to work very well.
The good news, though, was that the adaptive optics system was finally back up and running. The power had been restored after someone realized that one of the fans to cool the enormous power load of the laser system had not turned on after the initial power outage. A temperature sensor somewhere registered that things were heating up and shut everything down. This was good, otherwise all of the electronics would have been fried. But it was bad, because there was no obvious sign of what had happened except that nothing worked for a while. With some detective work and a few phone calls to people who thought they were going to have a free Saturday night to problem was found, the breaker switch was flipped, and the power came back on. The laser was still being massaged back to life by Kenny and company, but at least the adaptive optics system was going to work.
"OK, let's switch back" I said.
Part of Emily's Ph.D. thesis was the study of Titan with adaptive optics. Titan can be seen even though clouds, so at least this was going to work.
"The humidity is starting to spike" said Cindy's voice from the video screen.
Humidity. Argh. Telescope optics are delicate, so one does not want water condensing on them. If the humidity gets too high or if fog is around the telescopes must close immediately. I have been at the summit of Mauna Kea before on a beautiful clear night, walked outside, and in the slight moonlight been able to see that not a single telescope was open. Humidity. I'd rather that it just rained. Humidity is the worst. But as long as it stays low enough we were still ok.
"We are going to try not to think about the humidity and we're just going to go to Titan instead."
Jim chimed in, "We can stay open, but you should know that at moderately high humidity the lens in front of the laser has a tendency to fog up. We can then try to unfog the laser using the LLUD (Laser Lens Unfogging Device), but it doesn't work so well. If this happens we usually can't use the laser again for the rest of the night. We could for now install the LLCD (Laser lens covering device) that will keep the laser lens from fogging while we're not using it."
I asked what, exactly, where the LLUD and the LLCD. Jim answered, "The LLUD is a hair dryer. The LLCD is a large piece of plastic."
OK. If we have any hope of using the laser tonight we had better keep the lens dry. Let's do it.
Jim: "It requires pointing the telescope down towards the horizon and having someone stand up on the deck and place the piece of plastic up there. It'll take about 20 minutes."
Me & Emily: sigh.
Twenty minutes later, and about 3 hours after the sun set, we finally get to go to our first real target: Titan. But all of our staring at the satellite images had taught us one thing: the clouds were coming 3 hours after sunset. And our predictions were right. While Titan can be done pretty well though moderate clouds, we could barely see the thing.
Me & Emily: sigh.
After about 30 minutes there was word from Kenny: "The laser is ready to go!"
So at 11pm we were finally at full strength for the night, but the clouds covered the whole sky. We knew that the night was mostly lost, but we had hope that perhaps there would be a 30 minute sucker hole in the clouds that we could jump at. In just 30 minutes we could make a single observation of the positions of the moons of the Kuiper belt object 2003 EL61. Even one quick measurement would make us feel we had salvaged the night.
We swung the telescope to the position of 2003 EL61, watched the sky, and waited. Kenny kept the laser idling waiting to bring it up.
And we waited.
The satellite looked even more dismal than before.
A friend who was using the telescope in a few nights walked in to check on how we were doing. Our glum faces told the whole story.
"Do you mind if we use the telescope?" he asked.
"Can you do something useful in this mess?" I replied.
"Just maybe."
"Take it away. But we'll take it back if it ever clears."
Letting someone else take the telescope in really really bad conditions is a fabulous thing. Even with bad weather there is some residual guilt that I always feel about not taking data. Sometimes to assuage that guilt you take data that you know are worthless and that you will never use. But if someone else could possibly make use of the data all of your guilt is relieved.
My friend and his student then swung the telescope to the brightest star that they thought was interesting. The star was so bright that it could easily have been seen by eye if there were no clouds. But, now, the Keck telescope, the largest in the world, was having a difficult time even detecting it. Eventually the locked on and began to take data. We monitored how dim the star looked, and, knowing how bright it was supposed to be, could tell how bad the clouds were. In general the clouds made the star around 300 times dimmer than it is supposed to be. I suspect that they got no useful data throughout the night, either. But at least the guilt is now theirs.
By 3am I was tired and bored and the satellite image looked horrible. I gave up. I went to sleep.
Emily has more stamina so she decided to stay awake the last few hours. She promised to call if the sky ever cleared.
The phone never rang. The laser never fired. All of the people who sacrificed their Saturday nights to laser and adaptive optics were in turn sacrificed to the gods of weather. I apologized to and thanked everyone I knew who stayed and worked hard to make it happen for naught. No one seemed phased. Of course that's what they would do. Having the laser ready even on the slim chance that the clouds parted was the only thing that even occurred to them.
Astronomy is a pyramid, and, in this case, the pyramid is built on some pretty solid blocks.
People often ask: What happens if you have nights scheduled on the telescope and the weather conspires to prevent you from doing the project you had proposed to do? The answer: you can apply again next year. It is simply the luck of the draw.
We'll never know what we missed that night. Were the clouds on Titan doing something interesting (like the clouds on Earth were)? Where were the moons of 2003 EL61 that night? Could we have figured out what Orcus's moon was made of? We will apply again next year.
It's a long flight back home from Hawaii. Sometimes I sit on the airplane with my laptop salivating over all of the data that we collected. Sometimes I quietly meditate while thinking through the steps of the analysis that needs to be done. This time I slept. And I dreamt. And in my dreams the clouds parted, a bright yellow laser shot up into the sky, and the outer part of our solar system started to reveal its secrets.
