Designing and Scheduling Telescope Observations

Astronomers around the world submit proposals for time to observe with the Hubble and James Webb space telescopes. Ever wanted to know how the selected proposals are planned? Astronomer and planner Brigette Hesman and scientist Ryan Logue sat down for in-depth interviews to explain how a full year’s observations are broken up into small chunks and sent as telescope-readable files. Also read a behind-the-scenes account of NASA’s Double Asteroid Redirection Test (DART) impact from colleagues Tony Roman and Alison Vick. They share how they planned the observations of the fast-moving asteroid moonlet.

Planning the Long View, and Executing Day-to-Day Observations

For both long- and short-range planners, life centers on detailed timelines. For Hubble, creating observation schedules is mostly routine, though anomalies still periodically pop up that our staff address. But for Webb, the new “kid” on the block, the two planning branches have responded with care to every single detail. They’re learning its intricacies, and updating commands and procedures whenever its operating software is updated. Here, Brigette Hesman shares how her team plans for an upcoming year of observations and Ryan Logue explains how his team uses that information to build weekly schedules for both telescopes.

Brigette Hesman: Long-range planners for Hubble and Webb have to view telescope operations both week to week and in terms of a full year of programs. We are looking at long timescales and seeing how things fit into short timescales. We add flexibility to the schedules, because the more flexibility we provide, the more efficiently the telescopes run. And when the telescopes are efficient, everybody in the science community wins because we can take more observations. 

Our team provides short-term schedulers with a list of candidates for their weekly schedules, including a list of priorities for those visits. We call out programs that have to go within certain time periods, visits that are exceptionally long, or those that will generate a large volume of data. Webb has a lot of exoplanet programs, which require a lot more attention, since a planet passes in front of its star during a known but narrow time frame. We also allow time for all the station keeping and maintenance tasks, like downloading Webb’s data using the Deep Space Network. We spend a lot of time resolving potential conflicts and trying to make sure the schedule is seamless so short-term schedulers aren’t spending a lot of time iterating. 

I compare my work to what I do as a mom. My children attend school and activities throughout the week. It’s my job to make sure they have everything they need each day. But, like kids, there are times when the telescopes encounter issues. Stress can get pretty high, but that’s why we have a team. We support one another.

We initially build a long-term, one-year plan and then update it continuously in week-sized chunks as the cycle executes. We’re still learning a lot, including the nuances of the software, and fielding questions from scientists. Researchers who observe with Webb haven’t necessarily observed with Hubble, so they aren’t familiar with our processes. Plus, there are differences. Hubble has “failed” visits, but Webb has “skipped” visits. If a visit fails or is skipped, investigators need to work with us to get back on the schedule, but that doesn’t mean immediately. Researchers who have observed with Hubble understand, but it was new for people observing with Webb. Good communication from our mission office has helped fix that. Over time, Webb’s scheduling processes will become better known and understood by the research community. Even as we get to know Webb better, the demand for scheduling will still be high, because telescopes observe constantly. They don’t take weekends or holidays. 

Ryan Logue: I’m on the team that schedules the weekly calendars for both Hubble and Webb. We’re on a rotation. I do three weeks on Hubble, then three weeks on Webb. Every week, the long-range planners prep a giant, prioritized list of observations for us to schedule. For Webb, our focus is momentum management and data volume management, while maximizing time spent taking science. There’s only so much space on its recorders and we only have so many Deep Space Network contacts to get Webb’s data downlinked. We liken it to putting together a giant puzzle with no solution—and some of the pieces are tied together with string.

We make the most efficient calendars we can. We aim to maximize the time both telescopes spend taking science data. For Webb, we’re constantly working to reduce slew time. If one observation is in the southern hemisphere and another is in the northern hemisphere, instead of going from A immediately to B, we’ll stop at targets in between to minimize the slew time and maximize science. Once we have the short-term schedule finalized, we run software that turns our plans into spacecraft-readable files. We package those up and kick them over to the flight operations team members, who upload them to the telescope.

During the six months of commissioning, our workflow was a lot different. The mission planners on the flight operations team knew what needed to happen each day so that weeks later they could test another part of the instrument. Everything needed to happen in a precise order. That list would shift hour by hour, day by day, but we found a good cadence. Most of these observations were straightforward and we always had time to deal with issues that cropped up. We had an online huddle going constantly. We were also on the voice loops with the flight operations team in the Mission Operations Center. It all came down to communication and it meant our team of schedulers was never on our own.

Now, in normal operations, we’re focused on building seven-day schedules. We create a timeline and start ordering the observations, taking a week to build the next week’s timeline. As established as the processes are for Hubble and are getting for Webb, anomalies still crop up, whether with the telescope or the ground systems, and we respond. Our day-to-day work is very rarely routine. There’s always something new, every week, for either mission. 

Capturing Fast-Moving Targets with Webb and Hubble

Have you ever tried to observe a meteor shower? Look away and you might miss those streaks of light! Now, consider designing observations to capture fast-moving asteroids in the Solar System with Webb and Hubble. Both observatories are capable of locking onto nearby stars, known as guide stars, and snapping images before an asteroid or comet moves out of frame. Their observations are prepared by our deeply experienced technical staff. Here, two colleagues within the observation planning branch, Tony Roman and Alison Vick, recount how they helped capture the Double Asteroid Redirection Test (DART) impact—when a small NASA spacecraft collided with an asteroid moonlet known as Dimorphos—in late September.

Tony Roman: I help observers using Webb define the technical aspects of their observations and prepare the files to be sent to the telescope. I did this in partnership with a group of colleagues, which included fine guidance sensor engineers, short-term schedulers, scientists who specialize in moving target observations, and experts on Webb’s flight software. We knew a year or two in advance about this observation, but it sat on the backburner since we needed to launch and commission Webb first. 

The big issue with Webb was how fast the asteroid was moving: Webb was initially designed to track moving targets up to a rate of 30 milliarcseconds per second. This target was projected to be moving over 100 milliarcseconds per second. During Webb’s commissioning, we tested its moving target capabilities up to 67 milliarcseconds per second, which was more than double the design rate, and found that it worked fine. That built up our confidence. We thought, “Can we do 100?” We set up an engineering test to try. First, we searched the database of all known asteroids to find one that was moving at just the right speed. The limiting factor was the guide stars. Webb begins its observations by locking onto guide stars, which help the telescope point very finely. The engineering test was very difficult to put together, but we did it successfully, so we knew we could track the DART impact. 

We had to break one 45-minute observation into 12 so we could acquire new guide stars regularly and observe for a few minutes at a time. Of those, nine observations were completely successful, two were partially successful, and one failed. That was certainly more than good enough to get science data. Even after all that effort, the fact that Webb’s observations were not 100% successful shows that we were really were pushing the limits of what the telescope can do. 

Alison Vick: I’m also an expert on close, fast-moving targets. I studied comets in graduate school, which is how I ended up refining observations like this. The DART impact is one of the closest targets we’ve used Hubble to observe. Since the asteroid was moving too fast to rely exclusively on guide stars during the observations, we only used its gyros, spinning wheels that help stabilize and point Hubble, and can be used to navigate.

To design the observations, I had to account for parallax caused by the motion of Hubble around the Earth. Think about driving down the highway watching a road sign move against a background of trees. Just like the sign, the asteroid changes its position with respect to background stars. Then, there’s the motion of the asteroid. Combined, these movements create a curve, but Hubble only observes in straight lines. 

To accommodate this, we broke the observation into smaller chunks to mimic its motion. I went back and forth with the astronomer leading the observations to determine how much pointing error he could handle. We took one observation before the impact and six more over nine and a half hours. I watched the impact live on TV, too. I was so excited. 

Hubble got some usable data from every observation, despite the pointing difficulties and we were able to follow up a few weeks later, too. Hubble has been able to characterize these little orbiting piles of rock in ways I never thought possible. Every time I think we’ve reached the edge of what the telescope can do, studies show how we can push Hubble’s limits even further.