8 Teams Win Awards in 1st Year of Scialog: Early Science with the LSST
Top row: Wenbin Lu, Ana Bonaca, Kareem El-Badry, Allison Strom, Ben Margalit, Adam Miller, Caroline Morley. 2nd row: Igor Andreoni, Tanmoy Laskar, Mathew Madhavacheril, Alexander Ji, Vera Gluscevic, Anna Ho. 3rd row. Maya Fishbach, Elisabeth Newton, Charlotte Christensen, Nora Shipp, Burçin Mutlu-Pakdil, Krista Smith, Adi Foord.
Research Corporation for Science Advancement will make 21 separate awards of $60,000 in direct costs each to support the research of 20 scientists from colleges, universities, and research institutions in the United States and Canada in the first year of Scialog: Early Science with the LSST, a three-year initiative that aims to advance the foundational science needed to realize the full potential of the Vera C. Rubin Observatory’s upcoming Legacy Survey of Space and Time.
The Heising-Simons Foundation, The Brinson Foundation, the Leinweber Foundation, and independent philanthropist Kevin Wells are providing support to RCSA to fund the work of the eight cross-disciplinary teams.
The initiative represents a fulfilling new chapter in the story of RCSA’s long-term support of the Rubin Observatory, located in north-central Chile.
RCSA played a foundational role in its birth, helping establish the LSST Corporation (now the LSST Discovery Alliance) in 2003 and providing early financial and business support to launch the telescope's construction, including the design and construction of its primary mirror starting in 2007.
“It took more than 20 years to go from RCSA’s early commitment to the launching of this initiative,” said RCSA President & CEO Daniel Linzer. “After a lot of uncertainty over the years about the telescope and its construction, we are thrilled to see it nearing completion and look forward to great science coming from this Scialog that will help move astronomy and astrophysics forward for decades to come.”
Scialog is short for “science + dialog.” Created in 2010 by RCSA, the Scialog format aims to accelerate breakthroughs by building a creative network of scientists that crosses disciplinary silos, and by stimulating intensive conversation around a scientific theme of global importance.
The inaugural conference, held November 14-17 in Tucson, Arizona, engaged more than 50 observational astronomers, cosmologists, theoretical physicists and astrophysicists, computational modelers, data scientists, and software engineers from the U.S., Canada, and Chile in a series of conversations designed to build a networked community, discuss challenges and gaps in current knowledge, and form teams to propose blue-sky projects based on ideas generated during the conference.
The conference was also attended by representatives from the LSST Discovery Alliance, the Rubin Observatory, and other science-supporting foundations interested in Scialog’s unique process.
As the conference convened, Bob Blum, Rubin Observatory’s Director of Operations, discussed the recent successful use of the commissioning camera, which came online in October 2024.
“There's lots of challenges,” he said. “The system isn't reliable yet, but when it works, we're taking great data.”
With technical first light on the Rubin Observatory LSST Camera (the world’s largest digital camera) expected by early June 2025, full operations could start in September or October 2025. He said the first data preview should be available to researchers in March 2025, and the second in March 2026.
In time, the observatory will be able to survey the entire sky in only three nights and is expected to generate more than 20 terabytes of data each night, amassing a set of data and images that could address some of the deepest questions about the universe, its evolution, and the objects within it.
David W. Hogg, Professor of Physics and Data Science at New York University and Senior Research Scientist at the Flatiron Institute, provided a spark for discussions with his provocative keynote talk on the limitations and applications of machine learning in astrophysics, “How We Will (and How We Won’t) Use Machine Learning in the LSST Era.”
“Machine learning has become integrated into a lot of fields, but there's no really important result in astrophysics that can be attributed to machine learning,” he said. “While I’m going to say that machine learning methods cannot be trusted, I'm also going to say we can, and even must, use them in some instances. What I want us to do is get serious about what is good and use it.”
Hogg emphasized the importance of understanding the limitations of machine learning models, such as model misspecification and confirmation bias. He argued that while machine learning can be useful in outlier detection and causal situations, it should be used responsibly and with proper calibration.
Hogg said the problem of trust in machine learning is not just an issue for astrophysics but for all of society. He said working on improving accuracy of emulators and AI systems in astrophysics could serve as a "sandbox" that could have a "really, really big societal impact" by building broader trust in AI systems.
Jackie Faherty, a senior scientist and senior education manager at the American Museum of Natural History, generated excitement about the scientific potential of the LSST with her keynote talk, “The Stellar and Substellar Revolution with the Rubin Observatory.”
She called the LSST a "revolutionary new toy" that will provide a vast amount of data and open up new and unexpected avenues for exploration across various fields of astrophysics.
She said it will complement and build upon the discoveries made by previous surveys like Gaia, allowing it to fill in gaps and study objects that were previously difficult to observe. In addition, LSST's wide field of view, high cadence, and depth will enable it to make transformative discoveries, particularly in the study of low-mass stars, brown dwarfs, variable stars, and the structure of the Milky Way.
She said its “potential for serendipitous, interesting discovery” could lead to unexpected new information about rare and faint objects in the Milky Way, eclipsing binary systems, the galactic halo, and tracing the origins of interstellar debris to potential stellar hosts, among others.
She also said LSST’s huge data volume presents an opportunity to involve and engage the public in the process of discovery.
“It's a lot of data, but as somebody who works with the public every day, I know people are hungry to participate," she said.
An expert group of scientists served as Facilitators to guide discussions at the conference. Along with Hogg and Faherty, they included: Fred Adams, University of Michigan; Eric Bellm, University of Washington and LSST; Rebecca Bernstein, Giant Magellan Telescope; Lars Bildsten, Kavli Institute for Theoretical Physics; Xiaohui Fan, University of Arizona; Enrico Ramirez-Ruiz, University of California, Santa Cruz; Jeno Sokoloski, Columbia University and LSST; and Beth Willman, LSST Discovery Alliance.
Sokoloski also gave a presentation on accessing community resources such as LSST data-analysis tools and ways to participate in further developing resources.
During the course of the conference, Fellows met in a series of small- and medium-sized breakout groups to maximize interaction with other scientists with whom they have not collaborated. They were challenged to identify bottlenecks to progress in the field and brainstorm potential areas of inquiry where new, cross-disciplinary research might prove fruitful.
Participants are encouraged to get to know each other and their research, and to envision what it would look like to collaborate, how they could leverage their different approaches and methods, and what novel problems they could investigate together. On the final morning, teams that came together during the previous three days made brief proposal pitches for a total of 33 projects they had brainstormed at the conference.
The second meeting of Scialog: Early Science with the LSST is scheduled for November 13 – 16, 2025.
The following eight LSST teams will receive 2024 Scialog Collaborative Innovation Awards:
Wenbin Lu, Astronomy, University of California, Berkeley
Ana Bonaca, Carnegie Observatories, Carnegie Institution for Science
Kareem El-Badry, Astronomy, California Institute of Technology
IMBH in the LMC? A Hypervelocity Star Survey with LSST
Allison Strom, Physics and Astronomy, Northwestern University
Ben Margalit, Physics and Astronomy, University of Minnesota Twin Cities
Adam Miller, Physics and Astronomy, Northwestern University
Not So Heavy Metal: An Enhanced Rate of SLSNe at Cosmic Noon
Kareem El-Badry, Astronomy, California Institute of Technology
Caroline Morley, Astronomy, University of Texas at Austin
White Dwarf Companions as Brown Dwarf Chronometers
Igor Andreoni, Physics and Astronomy, University of North Carolina at Chapel Hill
Tanmoy Laskar, Physics & Astronomy, University of Utah
Mathew Madhavacheril, Physics and Astronomy, University of Pennsylvania
Rubin LSST as a Multi-Wavelength Discovery Engine for Relativistic Transients
Alexander Ji, Astronomy & Astrophysics, University of Chicago
Vera Gluscevic, Physics and Astronomy, University of Southern California
A Unified Model of Stellar Systems in LSST-Y1 for Dark Matter Inference
Anna Ho, Astronomy, Cornell University
Maya Fishbach, Canadian Institute for Theoretical Astrophysics, University of Toronto
Elisabeth Newton, Physics and Astronomy, Dartmouth College
Multimessenger Transients in AGN Disks
Charlotte Christensen, Physics, Grinnell College
Nora Shipp, Astronomy, University of Washington
Burçin Mutlu-Pakdil, Physics and Astronomy, Dartmouth College
Dwarf Debris and Dark Matter: Searching for Evidence of Hierarchical Formation in the Stellar Halos of Dwarf Galaxies
Krista Smith, Physics and Astronomy, Texas A&M University
Adi Foord, Physics, University of Maryland, Baltimore County
Towards a Census of Dual AGN Across Cosmic Time
