Waiting for Christmas
The space telescope that more than anything else defines my career as an astronomer is not JWST. It is Spitzer.
I began thinking about the Spitzer Space Telescope around 2001, when I started my PhD in Tautenburg observatory. At that point Spitzer was still SIRTF and JWST was still NGST. I wanted to use Spitzer to find out how brown dwarfs form, a question that came out of the stuff that I was obsessed with back then. In 2005, while a postdoc in Toronto, I applied for observing time with Spitzer, the most sensitive infrared telescope in the world at the time. I got twenty hours. For twenty hours, the newly launched space telescope looked exactly at the things that I found interesting. It followed my instructions like a well-behaved dog. While doing so, Spitzer took spectacular images like this one here:
Okay, it’s not that spectacular. But it’s a young brown dwarf in the middle, seen at an infrared wavelength of 24 microns, and it only appears bright because the brown dwarf is surrounded by some warm dust, a disc of warm dust. I remember the joy of planning the observations with the wonderfully simple Spitzer tools, the exhilaration of working with the clear, beautiful images, and the pleasure of writing papers that actually had convincing results. Brown dwarf have discs, some of these discs last 10 million years, and the discs contain relatively large dust grains. For a moment, science was so easy with Spitzer.
And then I was running out of momentum. In my second Spitzer proposal, I hit the wall, the hard, unrelenting sensitivity wall. The faintest young brown dwarfs, the ones with masses comparable to giant planets, were too faint for Spitzer. The error bars got inconveniently long. The conclusions got tentative. That was it. Another paper in 2008, and then it was done. The Spitzer party was over, for me.
This is the thing with the Next Big Telescope: You wait for it, years, sometimes decades. You work on it. You build it, some of us do. And you start using it. For a moment everything shines and glitters. Everything is easy, especially the things that seemed so hard just a moment ago. But because it is such a Great Telescope, you immediately design observations that push the telescope to its limits. And hence, just as the New Shiny Telescope is getting into the groove, you realise its limits. Sometimes even earlier. Sometimes before it’s even up there.
Around the time my honeymoon with Spitzer ended, I was starting to work on a survey for young brown dwarfs and free-floating planets, using the largest telescopes on Earth. This project was named SONYC, short for Substellar Objects in Nearby Young Cluster, ingeniously also the acronym of the String Orchestra of New York City. We found plenty of brown dwarfs, dozens, and a few free-floating objects with planetary masses. And this time we really had spectacular images, like this one.
At this point, we were still probing objects forming like stars, from the collapse of molecular clouds, for the most part at least. We were pretty sure about that. The next frontier was to find the real rogue planets, the ones ejected from their original homes, destined to roam the Galaxy without the stability and warmth of a host star in the vicinity. It was also clear that none of the existing telescopes could really find them, at least not if we are interested in having an image or a spectrum. We need something better, something bigger, something special.
And so, around ten years ago, I started working for the new telescope with the acronym JWST, the Just Wonderfully Superb Telescope, as it should be called, perhaps. In 2011 JWST was already twenty years in the making, already late, already too expensive. But the wait was worth it, always. JWST is the natural successor to Spitzer, but also the natural successor to WFCAM, the infrared instrument on-board of Hubble, and also the natural successor to the best infrared cameras on the largest ground-based telescopes. And it is the natural telescope to look for free-floating planets. That’s not my idea, smarter people had this thought a decade or two earlier.
It was very obvious that I am going to need that telescope. If I want to go from brown dwarfs to free-floating planets, from discs that may form planets around brown dwarfs to discs that may form moon around planets, I will need an awful lot of sensitivity. If I want to directly image the lowest mass objects that form like stars, or the most massive ejected planets, there is no alternative. JWST is the only game in town, at least for the rest of my career. The Juicy Wicked Space Tabernacle.
How do you even think about these big telescopes? How do you think about using a machine that took 30 years to make? That costs 10 billion dollars? That means, every hour of observing is going to be worth, like, 100000 pounds? Ugh? How do you wrap your head around this? How do you deal with the pressure, the stakes, the enormity? My sort of obvious advice is: You don’t. You approach it just like any other telescope. Because, in the end, that’s what it is. It’s just a thing that collects photons. You prepare your observations, you value every hour, you do your best, just like with an 80-cm telescope that is somewhere on a mountain in Spain. Eighty centimeter is, by the way, the smallest aperture telescope I have ever used for science. Of course JWST is larger and more sophisticated and in space, of course you are trying very hard not to make silly mistakes, just like with other telescopes. But there is also a lot more help. In contrast to the 80-cm in Spain, you can’t just start observing. Your program is going to be checked by an army of competent people. It’s going through years of probing and testing and discussing. Apart from that: Just another telescope.
So I twiddled my thumbs for a while, for a long while. I worked on stars, went back to unfinished projects, worked on datasets I had ignored for years. I worked with small telescopes, while thinking about the largest space telescope ever. And I tried to figure out how to get access to JWST.
Getting access to telescopes is one of the underrated parts of the job of an observational astronomer. I think of the entire set of telescopes worldwide as a toolbox, and the more tools I have, the more options I have to tackle problems. Access is critical. The default route to get access to a telescope is to write a good proposal that gets approved by my colleagues. But there are shortcuts, some more obvious than others. If you already have access to a few telescopes, you can sometimes leverage that to obtain time on other telescopes. You can occasionally swap time. You can use your existing datasets to produce synergies. Also, you can buy time, sometimes, but then you need to apply for money, not necessarily easier. You can also buy telescopes! In rare cases, you can ask nicely. And beyond that is the realm of guaranteed time.
Guaranteed time is awarded to people who have done a service to an observatory. The teams that built something, or wrote code, or simulated something, or developed a technology. In exchange, you get some telescope time. That’s how it works. Guaranteed time is primarily that, guaranteed. These observations will happen, as long as the telescope is working according to plan. But guaranteed time, being only available to selected few, is also less competitive than the regular proposals, where everybody can pitch in. You still need a good idea, a good target, and a good plan. But you might not have to convince all the rest of the world. Just some parts of it.
The obvious problem for me: I had done absolutely nothing for JWST. I just wanted to use it. But I had a science case, a good one, I thought, a nice target, and a clear plan that was buildings on years of work with Spitzer and SONYC. Not everybody has that. There is value in that, but it wasn’t clear to me how much. For a while I went around and shopped for guaranteed time. In other words, I was trying to get someone who has guaranteed time excited in free-floating planets.
It is not an accident, by the way, that I wasn’t interested in JWST earlier. It was a mission that was years and years away. As a postdoc on temporary contracts, I needed data now, not at some undefined moment years away. I’m not saying that it is impossible to build a research career while preparing a space mission that is ten years away, but it is difficult, at least psychologically. I used to think on timescales of a few months, a year at the most. What is the next paper, when is the next deadline, and what dataset is going to be important for my next job. For a postdoc, the time is now. The long-term future almost does not matter. It would have taken a lot more guts and a lot more commitment to a career in astronomy to work on JWST. Or on Gaia, or on LSST, or on any other future machinery, for that matter. A career that was by no means guaranteed. What do you want with guaranteed time on a space telescope ten years away, if your job is not guaranteed? Precarity breeds precarious thoughts. Only around 2012, when I had the first position that at least promised a long-term future, I started to get serious. At the time, that was only six years prior to launch. A blink of an eye.
With hindsight it is not entirely clear to me how I eventually ended up having 19 hours of JWST time. Or, rather, how I ended up leading a team that has 19 hours of JWST time. I do know that my trusted collaborators Kora and Ray were always there, every step of the way. It was a convoluted process that took several years. It involved a number of short trips to conferences and meetings, just for the purpose of getting the word out. It entailed talking to a lot of people, always about the same stuff, about free-floating planets. I got to talk about my science, over and over again. I had to rewrite the science case, over and over again. The allocation of guaranteed time is of course based on scientific arguments, but it is also a social process. In the end, which means around 2016, we were welcomed by the brilliant people from the Canadian NIRISS Guaranteed Time team — which became the home for this project.
This was not the end of the process. The program morphed from one iteration to the next. Every time the JWST tools changed, our program had to change. Every time the modeled observing sequences changed, our program had to be checked. Every hour was discussed, over and over again. Every minute is valuable. But it was clearly a converging process, and as time went on, the changes got smaller and smaller, until they were almost irrelevant. The final allocation, then another final allocation, and then another one. This one is really final. No, this one. Normal proposals have one deadline, guaranteed time proposals have a million.
I know exactly what I did when the launch was scrapped. It was late March 2018, I was in Dublin, writing a proposal for the first year of JWST, this time a regular proposal. The deadline was a few days away. The proposal was half finished, which means, I had a rough idea what I wanted to do, but didn’t know how. At that point, the launch was anticipated for October 2018. I probably found out over Twitter, because that’s how I find out about everything. The launch is postponed, at least for two years. The deadline for proposals is postponed as well. Two years is a long time in the world of brown dwarfs. I dropped all work immediately, walked over to the other office to tell Antonella, and then had a long coffee break. All of a sudden, I had a lot more spare time. Most of it was spent at the swimming place.
If it was only me, two years delay is not a big deal. I can wait. I have a permanent job! Really! I can sit this out. And the thing that caused the delay was not related at all to the parts of the machine that I need to worry about. It’s someone else’s job to fix all that. And since our program was already done and finished, it really just means finding something else to do for a couple of years. No problem, man.
The only downside of having a permanent job is, really, that I won’t have time to deal with the JWST data, once it actually arrives, whenever that is. Data analysis of that nature will take weeks or months of dedicated work, writing software, testing, rewriting software, and not just a few days here and there, an afternoon in between meetings, a couple of hours to procrastinate from marking. It will take a professional astronomer with some experience, who has all the time in the world, or rather, all the time of a postdoc contract.
In January 2017 I had applied for funding to hire someone a postdoc to prepare the data analysis, and then, after launch, to analyse the data. In November 2017 I found out that this proposal was successful. It was the first time I managed to get money for a postdoc. We shortlisted for this position in March 2018, just before I left for Dublin. Back then a lot of things happened in a single month! Unthinkable now. Times are slower, much slower. And now this.
The money, you see, does not care whether or not the telescope is actually in space. Although I got the money to work on JWST data, it is not going to wait for the telescope. It will start flowing in 2018, no matter what. And I will run out of funding in 2021, no matter what. So we interviewed, and we hired. In 2020 the continuation of the grant, basically the same proposal, was not successful. It was rejected, for one reason or the other. That puts me in a rather curious position: Right now, when the launch of JWST is imminent, is also the time when the funding is running out. The only funding I ever had to pay someone to do the actual work. Ironic, isn’t it.
This is, by the way, not an ideal scenario. But it’s a consequence of the way our system is built. Even the fact that I have to ask for funding for the data analysis does not make sense. I will have the observing time, and the dataset, and it’s very valuable, in scientific terms. Of course it needs to be analysed. The program was already approved, by numerous people, by the people that matter. In an ideal world (and in some way this is how the US system operates), there is funding for people, directly tied to the access to the telescope. You have observing time, excellent, here is some money to analyse the data. Because a telescope is useless without people. A dataset is useless without someone to look at it. We talk a lot about technology, about machines, telescopes, insights, results, fantastic images, incredible insights — but none of that happens without people. People is what matters in science.
Still, after all that. Am I excited about JWST? Of course. Because this is why I became an astronomer: To see something new, something that nobody else has seen before, and telescopes are my chosen method to achieve that. When JWST is finally in space, when it has turned from this gigantic machine that people can touch to a tiny dot that we can only see with a telescope, when it finally starts sending home data, we are going to see a lot of things nobody on Earth has seen before.
So this is one of these rare moments that links the 20-year-old me, the guy who was dreaming about space and large telescopes, with the guy I am today. It is a loophole in time. There have been a few of those, because I was very fortunate. But not that many. Certainly less than ten. Even better: This is a moment that I will share with 20-year old people in the present, the people who currently dream about space and large telescope. Because I really don’t do that anymore. I dream about, well, that’s another story.
And with any luck it is be a bit anti-climactic: The next generation is not going to understand what is so great about the new machine. When I started incorporating the Gaia data into undergraduate practicals, the reaction from the students was always a bit muted. A bit less enthusiastic than I expected it to be. For me, a parallax for a solar-mass star at thousands of lightyears away was a revolution. For them it was just that. And the errorbar is so large! (More observations needed.)
And the same is going to happen with JWST again, hopefully. Because it is happening now. It is just the thing we have today. It’s normal. And because we are immediately going to see faint little dots again, measurements with large errorbars, as a result of going for the limits. Astronomy is always going to be about faint little dots, no matter how big the telescope. JWST is going to be normal for students today, just like the VLT and the other ground-based 8-m telescopes, revolutionary in the late 90s, appeared normal to me. And that’s how it should be. What is currently a dream machine, the Next Big Telescope, will soon become just another tool, maybe a particularly powerful tool.
The next generation is going to dream about something even cooler.
