The Center for Astrophysics | Harvard & Smithsonian
The Center for Astrophysics | Harvard & Smithsonian
Support Our Science

Machine Learning

As astronomers build increasingly larger observatories capable of seeing more objects in the sky, the amount of data they collect has gone beyond what humans can analyze without help. Instead, researchers turn to teaching computers to sift through the data, identifying important patterns and connections that might otherwise be missed. This process is called machine learning, and it’s an essential aspect of modern astronomy at the Center for Astrophysics.

Our Work

Center for Astrophysics | Harvard & Smithsonian scientists and engineers use machine learning in many different ways:

  • Identifying and classifying objects and transient events within large surveys of the sky. Those surveys include current projects such as the Baryon Oscillation Spectroscopic Survey (BOSS) and upcoming observatories like the Large Synoptic Survey Telescope (LSST). LSST in particular will produce as much as 20 terabytes per night.
    A One-Percent Measure of Galaxies Half the Universe Away

  • Using machine learning to study day-to-day changes in the Sun, in an effort to understand solar weather. The space-based Solar Dynamics Observatory (SDO) collects 1.5 terabytes of data daily to ensure as detailed a picture as possible. Computer algorithms help sort this data and identify important features, such as solar flares.
    Solar Dynamics Observatory

  • Identifying exoplanets using data from the Kepler/K2 observatory. This space telescope has collected years of data on more than 150,000 stars to find the tiny flickers indicating the presence of planets. Machine learning helps separate signs of planets from other fluctuations in light form those stars, as well as identifying exoplanets that would be hard to spot otherwise.
    Artificial Intelligence, NASA Data Used to Discover Eighth Planet Circling Distant Star

  • Hunting for transient events like supernovas and the light-producing counterparts to gravitational wave discoveries. These events are some of the most powerful cosmic cataclysms, but they last only for a relatively short time before fading. That means identifying and studying them as quickly as possible. Machine learning helps by automating the process of finding these transient events within other data, including observations where finding supernovas wasn’t the original intent.
    Astronomers See Light Show Associated With Gravitational Waves

panorama of the sky seen by the Pan-STARRS1 telescope

This view of the sky seen by the Pan-STARRS1 telescope in Hawaii contains 3 billion astronomical sources, ranging from nearby asteroids to distant galaxies. Researchers use machine learning techniques to help analyze the 2 petabytes of data collected.

Credit: Danny Farrow, Pan-STARRS1 Science Consortium and Max Planck Institute for Extraterrestial Physics

A Universe of Data

Machine learning plays a huge role when cataloguing large numbers of anything, like galaxies in surveys of the whole sky. Computers can learn to identify and classify galaxy types, find transient events like supernovas, and pick out features in galaxy clusters. With thousands of potential sources from large surveys, it would take humans far too much time which could be better spent on other tasks. Machine learning shifts the effort from astronomers to computers, which excel at tedious detail-oriented tasks. Computer algorithms can also apply what we’ve learned from high-resolution observational data to improve lower-resolution images, allowing astronomers to construct what that object might look like through a more powerful telescope.

In addition, machine learning is essential for “time-domain astronomy”: looking for events that change during observation. Those include:

  • Hunts for exoplanets, which are planets orbiting other stars. When these worlds pass between their host stars and Earth, they block a small amount of light. Tracking the duration and amount of light provides information about the planet’s size and orbit. Several exoplanets have been identified using machine learning, including a few in multiple-planet systems, where the signals are hard for a human to distinguish.

  • Tracking changes in the light from stars. Some stars are extremely “active”, producing flares at unpredictable intervals. Others are variable, changing brightness as they expand and contract. Computers are suited to catch these variations, which can be subtle compared with the sheer amount of data needed to find them.

  • Studying patterns in the “weather” on our own Sun. Multiple observatories monitor the Sun literally all the time, which produces a vast amount of data to sift through. Machines are very good at finding important fluctuations in the Sun’s activity within that mass of information.

  • Finding and classifying supernovas. The explosions of stars and white dwarfs are random, unpredictable occurrences for all practical purposes. Catching these transient events requires sifting through data that’s often collected for different purposes, such as catalogs of galaxies.

  • Locating asteroids, comets, and other faint objects in the Solar System. These bodies show up as transients in other astronomical images. Computer searches are well suited to find them in large catalogs of data.

  • Using machine learning to classify objects in astronomical images. For example, in a telescope image of a galaxy or cluster of stars, the resolution might not be good enough to distinguish individual stars. However, computer processing can pick out light from different stars in the population, similar to the way low-resolution photos can be sharpened to reveal details that can’t be seen in the original.

Both current and future observatories regularly process many terabytes — trillions of bytes — of data. With that much information, it can be hard to tell what is important and what is not, with the importance depending on what scientific questions are being asked. Astronomers can teach computers to process the flood of data to pick out those important pieces.

Many current and future observatories, including NASA’s Transiting Exoplanet Survey Satellite (TESS), will bring in even more data useful to many areas of research. As a result, machine learning will become more important in the coming years, and CfA scientists are on the cutting edge of that development.

Share this Page
Big Questions
Science Fields
Divisions
See All Staff

Related News

Projects

AstroAI

Atomic and Molecular Physics, High Energy Astrophysics, Optical and Infrared Astronomy, Radio and Geoastronomy, Solar, Stellar, and Planetary Sciences, Theoretical Astrophysics, Harvard University Department of Astronomy, Science Education Department, Central Engineering, Director's Office, Chandra X-ray Center, Institute for Theoretical Atomic Molecular and Optical Physics, Institute for Theory and Computation
AstroAI is a institute dedicated to the development of artificial intelligence to enable next generation astrophysics at the Center for Astrophysics | Harvard & Smithsonian. Learn More
1

Seamless Astronomy

Seamless Astronomy brings together astronomers, computer scientists, information scientists, librarians and visualization experts involved in developing tools and systems to study and enable the next generation of online astronomical research.
2

Telescopes and Instruments

Giant Magellan Telescope

Is Earth unique, or are there other planets with life in the Milky Way? To answer this question and many others, astronomers need larger and more sensitive observatories than anything we currently have. For that reason, the Center for Astrophysics | Harvard & Smithsonian is collaborating with a number of other institutions around the world to create the Giant Magellan Telescope (GMT), currently under construction in Chile. The GMT will consist of seven large mirrors acting in concert as one giant telescope 80 feet across. That large size provides an unprecedented view of the sky and the ability to detect the chemical composition of exoplanet atmospheres. Like NASA’s Hubble Space Telescope, the GMT will be a powerful tool across the field of astronomy, providing insights into the formation of planets, the structure of galaxies, and the evolution of the universe itself.

Visit the GMT Website
1

Lynx X-Ray Observatory

The Lynx X-ray Observatory is a proposed X-ray space telescope, designed to be the next generation following NASA’s Chandra X-ray Observatory and the European Space Agency’s upcoming Athena Telescope. This observatory will be able to take images of a larger area of the sky in X-ray light with better resolution than any existing telescope, allowing astronomers to study the first supermassive black holes, the hot gas surrounding galaxies as they evolve, and radiation from the earliest stars. Researchers from the Center for Astrophysics | Harvard & Smithsonian and other universities are presenting Lynx to NASA for consideration in the 2020 Decadal Survey, with a proposed launch date in the mid-2030s.

Visit the Lynx X-Ray Observatory Website
2

Magellan Telescopes

The twin Magellan Telescopes in Chile are each 6.5 meter optical telescopes. These telescopes are both equipped with instruments to take images and spectra of light from a wide variety of astronomical sources, including exoplanet systems, star-forming regions, supernova remnants, and interacting galaxies. The Magellan Telescopes — named Baade and Clay —  are hosted at the Las Campanas Observatory and are operated by a consortium of institutions, including the Center for Astrophysics | Harvard & Smithsonian, Carnegie Institution for Science, University of Arizona, the University of Michigan, and the Massachusetts Institute of Technology. In addition, CfA scientists and engineers provided a wide field f/5 focal system (including secondary mirror and corrective optics), and a powerful astronomical camera for use at the Clay telescope.

Visit the Magellan Telescopes Website
3

MMT Observatory

The MMT Observatory is the premier visible-light and infrared telescope jointly managed by the Smithsonian Astrophysical Observatory, part of the Center for Astrophysics | Harvard & Smithsonian, and the University of Arizona. Located at the Fred Lawrence Whipple Observatory (FLWO) in southern Arizona, this 6.5-meter (21 foot) telescope is used to study objects across the field of astronomy, from the Solar System to distant galaxies. The MMT has provided a testbed for new telescope technologies developed by scientists and engineers at the CfA and the University of Arizona.

Visit the MMT Website
4

Spitzer Space Telescope

Infrared light is the key to studying a wide range of astronomical systems, including newborn stars and planets, dust in distant galaxies, and material swirling around black holes. The Spitzer Space Telescope was NASA’s orbiting infrared observatory, and part of NASA’s Great Observatories program, along with the Chandra X-ray Observatory and the Hubble Space Telescope. Scientists and engineers at the Center for Astrophysics | Harvard & Smithsonian led the design, testing, and construction of the Infrared Array Camera (IRAC). The Spitzer Space Telescope was launched in 2003 and retired in 2020.

Visit the Spitzer Space Telescope IRAC Page
5