From Argonne National Laboratory: “Stargazing with Computers: What Machine Learning Can Teach Us about the Cosmos”

Argonne Lab
News from From Argonne National Laboratory

February 18, 2020
Shannon Brescher Shea
shannon.shea@science.doe.gov

Gazing up at the night sky in a rural area, you’ll probably see the shining moon surrounded by stars. If you’re lucky, you might spot the furthest thing visible with the naked eye – the Andromeda galaxy.

Andromeda Galaxy Adam Evans

It’s the nearest neighbor to our galaxy, the Milky Way. But that’s just the tiniest fraction of what’s out there. When the Department of Energy’s (DOE) Legacy Survey of Space and Time (LSST) Camera at the National Science Foundation’s Vera Rubin Observatory turns on in 2022, it will take photos of 37 billion galaxies and stars over the course of a decade.

Fritz Zwicky discovered Dark Matter when observing the movement of the Coma Cluster., Vera Rubin a Woman in STEM denied the Nobel, did most of the work on Dark Matter.

Fritz Zwicky from http:// palomarskies.blogspot.com

Coma cluster via NASA/ESA Hubble

Astronomer Vera Rubin at the Lowell Observatory in 1965, worked on Dark Matter (The Carnegie Institution for Science)


Vera Rubin measuring spectra, worked on Dark Matter (Emilio Segre Visual Archives AIP SPL)


Vera Rubin, with Department of Terrestrial Magnetism (DTM) image tube spectrograph attached to the Kitt Peak 84-inch telescope, 1970. https://home.dtm.ciw.edu

The Vera C. Rubin Observatory currently under construction on the El Peñón peak at Cerro Pachón Chile, a 2,682-meter-high mountain in Coquimbo Region, in northern Chile, alongside the existing Gemini South and Southern Astrophysical Research Telescopes.

The Vera C. Rubin Observatory Data Journey, Illustration by Sandbox Studio, Chicago with Ana Kova

Dark Matter Research

Universe map Sloan Digital Sky Survey (SDSS) 2dF Galaxy Redshift Survey

Scientists studying the cosmic microwave background [CMB]hope to learn about more than just how the universe grew—it could also offer insight into dark matter, dark energy and the mass of the neutrino.

[caption id="attachment_73741" align="alignnone" width="632"] CMB per ESA/Planck

Dark matter cosmic web and the large-scale structure it forms The Millenium Simulation, V. Springel et al

Dark Matter Particle Explorer China

DEAP Dark Matter detector, The DEAP-3600, suspended in the SNOLAB deep in Sudbury’s Creighton Mine

LBNL LZ Dark Matter project at SURF, Lead, SD, USA


Inside the ADMX experiment hall at the University of Washington Credit Mark Stone U. of Washington. Axion Dark Matter Experiment

The output from this huge telescope will swamp researchers with data. In those 10 years, the Vera C Rubin Observatory Camera will take 2,000 photos for each patch of the Southern Sky it covers. Each image can have up to a million objects in it.

“In terms of the scale of the data, the amount of the data, the complexity of the data, they’re well beyond any of the current data sets we have,” said Rachel Mandelbaum, a professor at Carnegie Mellon University and LSST Dark Energy Science Collaboration spokesperson. “This opens up a huge amount of discovery space.”

Scientists aren’t building the LSST Camera to just take pretty pictures. They want to identify, categorize, and measure celestial objects that can reveal information about the very structure of the universe. Understanding dark energy and other cosmological mysteries requires data on supernovae and galaxies. Researchers may even find entirely new classes of objects.

“There are going to be some objects that we have never seen before because that is the point of new discovery,” said Renée Hložek, an assistant professor of astrophysics at the University of Toronto, who works with the LSST Dark Energy Science Collaboration. “We will find a bunch of what we call weirdos, or anomalies.”

The sheer volume and strangeness of the data will make it difficult to analyze. While a stargazer new to an area might go out in the field with a local expert, scientists don’t have such a guide to new pieces of the universe. So they’re making their own. More accurately, they’re making many different guides that can help them identify and categorize these objects. Astrophysicists supported by the DOE Office of Science are developing these guides in the form of computer models that rely on machine learning to examine the Vera C Rubin Observatory data. Machine learning is a process where a computer program learns over time about the relationships in a set of data.

Computer Programs that Learn

Processing data quickly is a must for scientists in the Dark Energy Science Collaboration.

Dark Energy Survey


Dark Energy Camera [DECam], built at FNAL


NOAO/CTIO Victor M Blanco 4m Telescope which houses the DECam at Cerro Tololo, Chile, housing DECam at an altitude of 7200 feet

Timeline of the Inflationary Universe WMAP

The Dark Energy Survey (DES) is an international, collaborative effort to map hundreds of millions of galaxies, detect thousands of supernovae, and find patterns of cosmic structure that will reveal the nature of the mysterious dark energy that is accelerating the expansion of our Universe. DES began searching the Southern skies on August 31, 2013.

According to Einstein’s theory of General Relativity, gravity should lead to a slowing of the cosmic expansion. Yet, in 1998, two teams of astronomers studying distant supernovae made the remarkable discovery that the expansion of the universe is speeding up. To explain cosmic acceleration, cosmologists are faced with two possibilities: either 70% of the universe exists in an exotic form, now called dark energy, that exhibits a gravitational force opposite to the attractive gravity of ordinary matter, or General Relativity must be replaced by a new theory of gravity on cosmic scales.

DES is designed to probe the origin of the accelerating universe and help uncover the nature of dark energy by measuring the 14-billion-year history of cosmic expansion with high precision. More than 400 scientists from over 25 institutions in the United States, Spain, the United Kingdom, Brazil, Germany, Switzerland, and Australia are working on the project. The collaboration built and is using an extremely sensitive 570-Megapixel digital camera, DECam, mounted on the Blanco 4-meter telescope at Cerro Tololo Inter-American Observatory, high in the Chilean Andes, to carry out the project.

Over six years (2013-2019), the DES collaboration used 758 nights of observation to carry out a deep, wide-area survey to record information from 300 million galaxies that are billions of light-years from Earth. The survey imaged 5000 square degrees of the southern sky in five optical filters to obtain detailed information about each galaxy. A fraction of the survey time is used to observe smaller patches of sky roughly once a week to discover and study thousands of supernovae and other astrophysical transients.

“There are going to be some objects that we have never seen before because that is the point of new discovery,” said Renée Hložek, an assistant professor of astrophysics at the University of Toronto, who works with the LSST Dark Energy Science Collaboration. “We will find a bunch of what we call weirdos, or anomalies.”

The sheer volume and strangeness of the data will make it difficult to analyze. While a stargazer new to an area might go out in the field with a local expert, scientists don’t have such a guide to new pieces of the universe. So they’re making their own. More accurately, they’re making many different guides that can help them identify and categorize these objects. Astrophysicists supported by the DOE Office of Science are developing these guides in the form of computer models that rely on machine learning to examine the LSST data. Machine learning is a process where a computer program learns over time about the relationships in a set of data.

Computer Programs that Learn

Processing data quickly is a must for scientists in the Dark Energy Science Collaboration. Scientists need to know that the camera is pointing at exactly the right place and taking data correctly each time. This quick processing also helps them know if anything has changed in that part of the sky since the last time they took photos of it. Subtracting the current photo from previous ones shows them if there’s a sign of an interesting celestial object or phenomenon.

They also need to combine a lot of photos together in a way that’s accurate and usable. This project is looking into the depths of the universe to capture images of some of the faintest stars and galaxies. It will also be taking photos in less-than-ideal atmospheric conditions. To compensate, scientists need programs that can combine images together to improve clarity.

Machine learning can tackle these challenges in addition to handling the sheer amount of data. As these programs analyze more data, the more accurate they become. Just like a person learning to identify a constellation, they gain better judgement over time.

“Many scientists regard machine learning as the most promising option for classifying sources based on photometric measurements (measurements of light intensity),” said Eve Kovacs, a physicist at DOE’s Argonne National Laboratory.

But machine learning programs need to teach themselves before they can tackle a pile of new data. There are two main ways to “train” a machine learning program: unsupervised and supervised.

Unsupervised machine learning is like someone teaching themselves about stars from just their nightly observations. The program trains itself on unlabeled data. While unsupervised machine learning can group images and identify outliers, it can’t categorize them without a guidebook of some sort.

Supervised machine learning is like a newbie relying on a guidebook. The researchers feed it a massive set of data that is labeled with the classes of each object. By examining the data over and over, the program learns the relationship between the observation and the labels. This technique is especially useful for classifying objects into known groups.

In some cases, the researchers also feed the program a specific set of features to look for, like brightness, shape, or color. They provide guidance on how important each feature is compared to the others. In other programs, the machine learning program figures out the relevant features itself.

However, the accuracy of supervised machine learning depends on having a good training set, with all of the diversity and variability of a real one. For photos from the LSST Camera, that variability could include streaks from satellites moving across the sky. The labeling also has to be extremely accurate.

“We have to put as much physics as we can into the training sets,” said Mandelbaum. “It doesn’t remove from us the burden to understand the physics. It just moves it into a different part of the problem.”

See the full article here .

five-ways-keep-your-child-safe-school-shootings

Please help promote STEM in your local schools.

Stem Education Coalition

Argonne National Laboratory seeks solutions to pressing national problems in science and technology. The nation’s first national laboratory, Argonne conducts leading-edge basic and applied scientific research in virtually every scientific discipline. Argonne researchers work closely with researchers from hundreds of companies, universities, and federal, state and municipal agencies to help them solve their specific problems, advance America’s scientific leadership and prepare the nation for a better future. With employees from more than 60 nations, Argonne is managed by UChicago Argonne, LLC for the U.S. Department of Energy’s Office of Science. For more visit http://www.anl.gov.

The Advanced Photon Source at Argonne National Laboratory is one of five national synchrotron radiation light sources supported by the U.S. Department of Energy’s Office of Science to carry out applied and basic research to understand, predict, and ultimately control matter and energy at the electronic, atomic, and molecular levels, provide the foundations for new energy technologies, and support DOE missions in energy, environment, and national security. To learn more about the Office of Science X-ray user facilities, visit http://science.energy.gov/user-facilities/basic-energy-sciences/.

Argonne is managed by UChicago Argonne, LLC for the U.S. Department of Energy’s Office of Science

Argonne Lab Campus