Webb Observes a Globular Cluster Sparkling with Separate Stars

A rectangular image oriented horizontally appears to be two separate square images with a wide black gap in between. The two squares do not mirror each other exactly or align perfectly together. It looks instead like they are two parts of a larger image that has been obscured in the middle by black strip. Both squares are filled with blue, white, yellow, and red points of light of different size and brightness, most of which are stars. The larger and brighter stars show Webb’s distinctive diffraction pattern consisting of eight spikes radiating from the center. Both squares show an increase in density of stars toward the central gap. Altogether, the stars appear to form a loose ball-like shape whose core is obscured by the gap.
Image of the globular cluster M92 captured by the James Webb Space Telescope’s NIRCam instrument. This image is composite of four exposures using four different filters: F090W (0.9 microns) is shown in blue; F150W (1.5 microns) in cyan; F277W (2.77 microns) in yellow; and F444W (4.44 microns) in red. The black strip in the center is a chip gap, the result of the separation between NIRCam’s two long-wavelength detectors. The gap covers the dense center of the cluster, which is too bright to capture at the same time as the fainter, less dense outskirts of the cluster. The image is about 5 arcminutes (39 light-years) across. Download the full-resolution image of M92 from the Resource Gallery at the Space Telescope Science Institute. Image credit: NASA, ESA, CSA, A. Pagan (STScI).

Editor’s Note: This post highlights data from Webb science in progress, which has not yet been through the peer-review process.

On June 20, 2022, the James Webb Space Telescope spent just over one hour staring at Messier 92 (M92), a globular cluster 27,000 light-years away in the Milky Way halo. The observation – among the very first science observations undertaken by Webb – is part of Early Release Science (ERS) program 1334, one of 13 ERS programs designed to help astronomers understand how to use Webb and make the most of its scientific capabilities.

We spoke with Matteo Correnti from the Italian Space Agency; Alessandro Savino from the University of California, Berkeley; Roger Cohen from Rutgers University; and Andy Dolphin from Raytheon Technologies to find out more about Webb’s observations of M92 and how the team is using the data to help other astronomers. (Last November, Kristen McQuinn talked with us about her work on the dwarf galaxy WLM, which is also part of this program.)

Tell us about this ERS program. What are you trying to accomplish?

Alessandro Savino: This particular program is focused on resolved stellar populations. These are large groups of stars like M92 that are very nearby – close enough that Webb can single out the individual stars in the system. Scientifically, observations like these are very exciting because it is from our cosmic neighborhood that we learn a lot of the physics of stars and galaxies that we can translate to objects that we see much farther away.

Matteo Correnti: We’re also trying to understand the telescope better. This project has been instrumental for improving the calibration (making sure all of the measurements are as accurate as possible), for improving the data for other astronomers and other similar projects.

Why did you decide to look at M92 in particular?

Savino: Globular clusters like M92 are very important for our understanding of stellar evolution. For decades they have been a primary benchmark for understanding how stars work, how stars evolve. M92 is a classic globular cluster. It’s close by; we understand it relatively well; it’s one of our references in studies of stellar evolution and stellar systems.

Correnti: Another reason M92 is important is because it is one of the oldest globular clusters in the Milky Way, if not the oldest one. We think M92 is between 12 and 13 billion years old. It contains some of the oldest stars that we can find, or at least that we can resolve and characterize well. We can use nearby clusters like this as tracers of the very ancient universe.

Roger Cohen: We also chose M92 because it is very dense: There are a lot of stars packed together very closely. (The center of the cluster is thousands of times denser than the region around the Sun.) Looking at M92 allows us to test how Webb performs in this particular regime, where we need to make measurements of stars that are very close together.

What are the characteristics of a globular cluster that make it useful for studying how stars evolve?

Andy Dolphin: One of the main things is that the bulk of the stars in M92 would have formed at roughly the same time and with roughly the same mix of elements, but with a wide range of masses. So we can get a really good survey of this particular population of stars.

Savino: Also, since the stars all belong to the same object (the same globular cluster, M92), we know they are all about the same distance away from us. That helps us a lot because we know that differences in brightness between the different stars must be intrinsic, instead of just related to how far away they are. It makes the comparison with models much, much easier.

This star cluster has already been studied with the Hubble Space Telescope and other telescopes. What can we see with Webb that we have not seen already?

Cohen: One of the important differences between Webb and Hubble is that Webb operates at longer wavelengths, where very cool, low-mass stars give off most of their light. Webb is well-designed to observe very cool stars. We were actually able to reach down to the lowest mass stars – stars less than 0.1 times the mass of the Sun. This is interesting because this is very close to the boundary where stars stop being stars. (Below this boundary are brown dwarfs, which are so low-mass that they’re not able to ignite hydrogen in their cores.)

Correnti: Webb is also a lot faster. To see the very faint low-mass stars with Hubble, you need hundreds of hours of telescope time. With Webb, it takes just a few hours.

Cohen: These observations weren’t actually designed to push very hard on the limits of the telescope. So it’s very encouraging to see that we were still able to detect such small, faint stars without trying really, really hard.

What’s so interesting about these low-mass stars?

Savino: First of all, they are the most numerous stars in the universe. Second, from a theoretical point of view, they are very interesting because they’ve always been very difficult to observe and characterize. Especially stars less than half the mass of the Sun, where our current understanding of stellar models is a little more uncertain.

Correnti: Studying the light these low-mass stars emit can also help us better constrain the age of the globular cluster. That helps us better understand when different parts of the Milky Way (like the halo, where M92 is located) formed. And that has implications for our understanding of cosmic history.

It looks like there’s big gap in the middle of the image you captured. What is that and why is it there?

Dolphin: This image was made using Webb’s Near-Infrared Camera (NIRCam). NIRCam has two modules, with a “chip gap” between the two. The center of the cluster is extremely crowded, extremely bright. So that would have limited the usefulness of the data from that region. The position of these images overlaps nicely with Hubble data available already.

Square image filled with blue, white, yellow, and red points of light of different size and brightness, most of which are stars. The larger and brighter stars show Webb’s distinctive diffraction pattern consisting of eight spikes radiating from the center. At the lower right is a scale bar labeled 2 light-years. The scale bar is two-ninths the width of the image, and shows that throughout the image, the distance between adjacent stars is a fraction of one light-year. The density of stars and brightness of the image is greatest in the upper left portion of the image where the stars are much closer together, and decreases gradually toward the bottom right where they are farther apart. The number of larger, brighter stars also appears to decrease from the upper left toward the lower right.
Detail of the globular cluster M92 captured by Webb’s NIRCam instrument. This field of view covers the lower left quarter of the right half of the full image. Globular clusters are dense masses of tightly packed stars that all formed around the same time. In M92, there are about 300,000 stars packed into a ball about 100 light-years across. The night sky of a planet in the middle of M92 would shine with thousands of stars that appear thousands of times brighter than those in our own sky. The image shows stars at different distances from the center, which helps astronomers understand the motion of stars in the cluster, and the physics of that motion. Download the image detail of M92 from the Space Telescope Science Institute. Image credit: NASA, ESA, CSA, A. Pagan (STScI).

One of your main goals was to provide tools for other scientists. What are you particularly excited about?

Dolphin: One of the key resources we developed and have made available to the astronomical community is something called the DOLPHOT NIRCam module. This works with an existing piece of software used to automatically detect and measure the brightness of stars and other unresolved objects (things with a star-like appearance). This was developed for cameras on Hubble. Adding this module for NIRCam (as well as one for NIRISS, another of Webb’s instruments) allows astronomers the same analysis procedure they know from Hubble, with the additional benefit of now being able to analyze Hubble and Webb data in a single pass to get combined-telescope star catalogs.

Savino: This is a really big community service component. It’s helpful for everyone. It’s making analysis much easier.

About the Authors:

        • Matteo Correnti is a research fellow at the Space Science Data Center at the Italian Space Agency and National Institute of Astrophysics in Rome, Italy.
        • Alessandro Savino is a postdoc at the University of California, Berkeley.
        • Roger Cohen is a postdoc at Rutgers University in New Brunswick, New Jersey.
        • Andy Dolphin is a technical fellow at Raytheon Technologies in Tucson, Arizona.

Related Links:




Breaking the Tracking Speed Limit With Webb

In September, the James Webb Space Telescope observed as NASA’s Double Asteroid Redirection Test (DART) intentionally smashed a spacecraft into a small asteroid, in the world’s first-ever in-space test for planetary defense. Today, we hear from Stefanie Milam, Webb’s deputy project scientist for planetary science at NASA’s Goddard Space Flight Center, about how the Webb team worked to capture these one-of-a-kind observations.

To observe NASA’s DART mission impact event, Webb had to exceed its required and tested observatory capabilities. Before launch, astronomers had plans to observe solar system objects with Webb, which meant it was designed to track targets that move with respect to stars and galaxies in the distant universe. With Webb being a large boat-shaped observatory with its large sunshield and prominent mirror standing on top resembling a sailboat, this was considered a challenge and something we wanted to both simulate before launch, as well as test rigorously during commissioning. We had to make sure that we could track these moving objects with the same precision that we have when pointing at fixed targets. Generally, using a special camera called the Fine Guidance Sensor (FGS), Webb locks onto a so-called guide star to stay pointed at its target with great precision. Moving target observations are executed by moving the guide star at the precise rate of the moving object inside the FGS field of view. This ensures the science target remains stationary in the specified science instrument. This of course also means the background field of stars and galaxies drift in the moving target images, not unlike taking a picture of a race car while turning the background spectators into a streaky blur.

Series of Webb NIRCam observations (filter: F070W) from just before DART impact.
at 7:14 p.m. EDT, Sept. 26, through five hours post-impact. The success of these observations was dependent on implementation of faster moving-target tracking rates. Image Credits: Science: NASA, ESA, CSA, Cristina Thomas (Northern Arizona University), Ian Wong (NASA-GSFC); Joseph DePasquale (STScI)

We worked diligently with the FGS team and the observatory flight software engineers to ensure that pre-flight simulations of the observatory demonstrated a tracking capability with a speed limit originally set at the rate of Mars (30 milliarcseconds per second or the width of a full moon in just under 17 hours!). Webb launched with this notional speed limit. But solar system scientists – especially those who study fast-moving small bodies like asteroids, comets, and interstellar objects – really wanted to observe objects that moved faster than Mars. Of course, we could observe these objects when they are further away from Earth and the Sun and therefore moving slower, but that is not always aligned with when they are the most interesting, e.g., when a comet undergoes an outburst.

So, we set out to not only confirm that Webb could track at the pre-flight speed limit during commissioning, but also to show that the observatory is capable of much faster tracking rates. We started out slow with our first asteroid, 6481 Tenzing, at a rate of only 5 milliarcseconds per second (mas/s) (or 18 for other asteroids. All of these tests were successful, and we confirmed that Webb was really good at tracking moving objects at a a rate of up to 67 mas/s (241ʺ/hr).

This was great news for all solar system scientists planning to use Webb!

Our next challenge was upon us when the Guaranteed Time Observation Program to study Didymos during the DART mission impact (program 1245, PI: Cristina Thomas at Northern Arizona University) was being planned. Program coordinator, Tony Roman at the Space Telescope Science Institute wanted to try to track the asteroid with Webb during the impact of the DART spacecraft (at over 100 mas/s or 360ʺ/hr) with Webb. According to Roman, “Tracking at a rate slower than Didymos’ actual motion would have resulted in blurred images. If we could track at Didymos’ actual rate, that would make the most of this unique opportunity.” This was far faster than our fastest rates from commissioning, but the project was convinced we should try this in order to support another NASA mission, in conjunction with simultaneous observations with the Hubble Space Telescope. Tests were set up on the flight simulator to verify the software and hardware could handle these super-fast tracking rates. The system was tweaked several times to optimize the performance, and to ensure that we could track an object moving that fast. Once convinced by the simulator results, we planned some engineering tests on the spacecraft with two observations of near-Earth asteroid 2010 DF1 at rates of motion that were (324 and 396ʺ/hr).  This was the fastest and brightest asteroid in Webb’s field of regard we could test on near the DART impact speed.  If we could successfully track this asteroid, we knew we could track the DART impact event. We were down to the wire with tests being done only two weeks prior to the date of the DART impact into Dimorphos!

Series of observations from Webb’s Near-Infrared Camera (NIRCam) tracking asteroid 2010 DF1 at 90 mas/s (Filter: F322W2). This is 10 images from Sept. 9, with a total exposure duration of 311 seconds. Image credit: Ian Wong (NASA-GSFC).

Observations to monitor the DART impact were prepared and uplinked to the observatory; at the scheduled time, just prior to the impact, the series of observations began executing and were successful. With data spilling into the archive for analysis, Ian Wong of NASA Goddard completed a quick, but thorough, analysis, which verified the super-fast rates were successful! We had demonstrated that we could plan for an event like the DART impact with Webb, and had beautiful results to show. While the analysis of the impact data , we have now confirmed that we can track and observe targets over 100 mas/s (360ʺ/hr) with Webb. However, we will likely not use the high rates routinely. While successful, they were challenging to plan and schedule. Guide stars only stay within the FGS field of view for a short time at those speeds. That means we would have to use multiple guide stars to support a longer observation, and changing guide stars introduces complexity and inefficiency. In the end, the new speed limit set for Webb is now 75 mas/s for future observations, but special permission may be requested for rates up to 100 mas/s.

– Stefanie Milam, Webb’s deputy project scientist for planetary science, NASA Goddard