Science1 publisher3 min readPublished
Rubin Observatory aimed its first research image and catalog at the well-studied COSMOS deep field
The LSST Camera's first image released for science stacks hundreds of exposures of the COSMOS field, where Hubble worked two decades ago, and Rubin's deputy director calls the field a testing ground for the ten-year survey.
The Scientist · Science desk
What happened
- Rubin Observatory's LSST Camera has produced an image and catalog for scientific research for the first time, and the target was the COSMOS deep field in the constellation Sextans.
- The image was built by stacking hundreds of individual observations from the 3.2-gigapixel camera rather than from any single long exposure.
- Hubble imaged the same region in segments from 2003 to 2005 with its Advanced Camera for Surveys, and the frames were later stitched into a single mosaic.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- capability Groups that have written analysis code against simulated Rubin outputs can now run it on a real catalog covering a field where two decades of independent measurements already exist to check against.
- constraint A single deep stack does not test the part of the survey Rubin was designed around, and Blum ties the transient and variable-object demonstration to repeat visits over the next few years.
- decision Anyone weighing whether to move a deep-field program onto Rubin data has no published depth figure to compare against Hubble-era catalogs, so for now the choice rests on the pipeline's behaviour.
- precedent Choosing a calibration field for the first science release sets the terms for the early literature: disagreements with the prior record will be read as instrument checks before they are read as discoveries.
Pointing a new instrument at the best-studied patch of sky is a calibration decision. COSMOS has been observed for more than two decades in optical, radio, ultraviolet and X-ray light [4]. That history is why the field has been used as a reference for testing new instruments, comparing measurements and combining observations, phys.org reports [5]. Where Rubin's catalog disagrees with the existing record for a galaxy in that field, the first thing to check is the pipeline.
The released image is a stack of hundreds of individual exposures from the 3.2-gigapixel LSST Camera [6], and it holds more than half a million galaxies against more than 50,000 stars [7], about ten galaxies for every star [17]. That balance follows from where the telescope was pointed. The camera looked away from the crowded plane of the Milky Way, so the galactic disk and its clouds of dust and gas did not obscure the field [10]. The stars that do appear are Milky Way stars along the line of sight [9]. The frame includes barred spirals with arms, smooth ellipticals, faint red galaxies and distorted merging galaxies [8].
Phil Marshall, deputy director of Rubin Observatory at SLAC National Accelerator Laboratory, said the field's observing history is what makes it useful now. "Its wealth of prior observations, and its repeated targeting both during commissioning and as one of LSST's deep fields," he said, "will make it very valuable as a testing ground for scientists as they get ready to take on the survey data" [12].
A stack exercises one part of the survey design: combining exposures and extracting a catalog from them. The variable and explosive sky is a separate demonstration, and Bob Blum, director of Rubin Observatory at NSF NOIRLab, put it in the future tense. "Repeated visits to the field over the next few years will demonstrate the power of our survey design for discovery," Blum said, "by providing our science community with a huge number of transient and variable objects like supernovae and other explosive transients for follow-up and detailed study" [14].
The phys.org account does not include a limiting magnitude or compare Rubin's counts against the existing COSMOS catalogs [16]. Hubble imaged the region in segments from 2003 to 2005 with its Advanced Camera for Surveys [2]; Rubin's science-validation data begin in April 2025 [13], twenty years later [18].
Early Data Preview 2 combines Rubin observations taken between April 2025 and January 2026 [13], a window of about nine months [19]. It is the first phase of Data Preview 2, and Rubin describes it as a preview of the ten-year survey that also lets scientists test and validate the observatory's tools and data products [15].
What to watch
- Whether the full Data Preview 2 release publishes limiting magnitudes and photometric comparisons against existing COSMOS catalogs.
- The first difference-imaging results from the repeat COSMOS visits Blum places over the next few years.
- Whether independent groups reproduce Hubble-era COSMOS measurements in Rubin's catalog, and where the two disagree.