Tag Archives: #supernovae

Artificial Intelligence Classifies Supernova Explosions With Unprecedented Accuracy (Astronomy)

A new machine learning algorithm trained only with real data has classified over 2,300 supernovae with over 80% accuracy.

Artificial intelligence is classifying real supernova explosions without the traditional use of spectra, thanks to a team of astronomers at the Center for Astrophysics | Harvard & Smithsonian. The complete data sets and resulting classifications are publicly available for open use.

Cassiopeia A, or Cas A, is a supernova remnant located 10,000 light years away in the constellation Cassiopeia, and is the remnant of a once massive star that died in a violent explosion roughly 340 years ago. This image layers infrared, visible, and X-ray data to reveal filamentary structures of dust and gas. Cas A is amongst the 10-percent of supernovae that scientists are able to study closely. CfA’s new machine learning project will help to classify thousands, and eventually millions, of potentially interesting supernovae that may otherwise never be studied. Credit: NASA/JPL-Caltech/STScI/CXC/SAO

By training a machine learning model to categorize supernovae based on their visible characteristics, the astronomers were able to classify real data from the Pan-STARRS1 Medium Deep Survey for 2,315 supernovae with an accuracy rate of 82-percent without the use of spectra.

The astronomers developed a software program that classifies different types of supernovae based on their light curves, or how their brightness changes over time. “We have approximately 2,500 supernovae with light curves from the Pan-STARRS1 Medium Deep Survey, and of those, 500 supernovae with spectra that can be used for classification,” said Griffin Hosseinzadeh, a postdoctoral researcher at the CfA and lead author on the first of two papers published in The Astrophysical Journal. “We trained the classifier using those 500 supernovae to classify the remaining supernovae where we were not able to observe the spectrum.”

Edo Berger, an astronomer at the CfA explained that by asking the artificial intelligence to answer specific questions, the results become increasingly more accurate. “The machine learning looks for a correlation with the original 500 spectroscopic labels. We ask it to compare the supernovae in different categories: color, rate of evolution, or brightness. By feeding it real existing knowledge, it leads to the highest accuracy, between 80- and 90-percent.”

Although this is not the first machine learning project for supernovae classification, it is the first time that astronomers have had access to a real data set large enough to train an artificial intelligence-based supernovae classifier, making it possible to create machine learning algorithms without the use of simulations.

“If you make a simulated light curve, it means you are making an assumption about what supernovae will look like, and your classifier will then learn those assumptions as well,” said Hosseinzadeh. “Nature will always throw some additional complications in that you did not account for, meaning that your classifier will not do as well on real data as it did on simulated data. Because we used real data to train our classifiers, it means our measured accuracy is probably more representative of how our classifiers will perform on other surveys.” As the classifier categorizes the supernovae, said Berger, “We will be able to study them both in retrospect and in real-time to pick out the most interesting events for detailed follow up. We will use the algorithm to help us pick out the needles and also to look at the haystack.”

The project has implications not only for archival data, but also for data that will be collected by future telescopes. The Vera C. Rubin Observatory is expected to go online in 2023, and will lead to the discovery of millions of new supernovae each year. This presents both opportunities and challenges for astrophysicists, where limited telescope time leads to limited spectral classifications.

“When the Rubin Observatory goes online it will increase our discovery rate of supernovae by 100-fold, but our spectroscopic resources will not increase,” said Ashley Villar, a Simons Junior Fellow at Columbia University and lead author on the second of the two papers, adding that while roughly 10,000 supernovae are currently discovered each year, scientists only take spectra of about 10-percent of those objects. “If this holds true, it means that only 0.1-percent of supernovae discovered by the Rubin Observatory each year will get a spectroscopic label. The remaining 99.9-percent of data will be unusable without methods like ours.”

Unlike past efforts, where data sets and classifications have been available to only a limited number of astronomers, the data sets from the new machine learning algorithm will be made publicly available. The astronomers have created easy-to-use, accessible software, and also released all of the data from Pan-STARRS1 Medium Deep Survey along with the new classifications for use in other projects. Hosseinzadeh said, “It was really important to us that these projects be useful for the entire supernova community, not just for our group. There are so many projects that can be done with these data that we could never do them all ourselves.” Berger added, “These projects are open data for open science.”

This project was funded in part by a grant from the National Science Foundation (NSF) and the Harvard Data Science Initiative (HDSI).

References: (1) SuperRAENN: A semi-supervised supernova photometric classification pipeline trained on Pan-STARRS1 Medium Deep Survey supernovae, A. Villar et al, The Astrophysical Journal, 2020 December 17
[doi: 10.3847/1538-4357/abc6fd, preprint: https://arxiv.org/pdf/2008.04921.pdf] (2) Photometric Classification of 2315 Pan-STARRS1 Supernovae with Superphot by G. Hosseinzadeh et al, The Astrophysical Journal, 2020 December 17,
[doi: 10.3847/1538-4357/abc42b, preprint: https://arxiv.org/pdf/2008.04912.pdf]

Provided by Harvard Smithsonian Center for Astrophysics

Neil Gehrels Swift Observatory Gamma-Ray Burst Associated with Kilonovae: Ambushing the Standard Candle in its own Nest (Astronomy)

Gamma-Ray Bursts (GRBs) are the most luminous and explosive transient phenomena in the Universe after the Big Bang, but they are still puzzling phenomena regarding their emission mechanism even after more than 50 years from their discovery. A powerful tool for characterizing and classifying GRBs to allow them to be used as tracers of the expansion history of the Universe and to understand their mysterious and debated physical mechanisms has been recently presented by an international team of researchers led by Dr. hab. Maria Dainotti, Assistant Professor at Jagiellonian University, Poland and concurrently serving as Senior Research Scientist at RIKEN and affiliate Research Scientist at Space Science Institute, in Boulder, Colorado.

Illustration 1: NASA’s Swift spacecraft spots its thousandth gamma-ray burst. Credit: NASA.

The new article, which has been accepted by the Astrophysical Journal, pays particular attention to the GRBs associated with Kilonovae, and to a sample called the Platinum sample for which the maximum redshift observed is 5, much more distant than the maximum redshift at which the SNe Ia have been observed.

Astronomers can only directly measure distances to objects that are close to Earth and can extrapolate the distances to objects farther out. All the objects that serve as rungs on the cosmological distance ladder have known luminosities and are referred to as “standard candles”. Once the absolute luminosity of the standard candle is known, the distance to that object can be calculated based on its measured brightness. For example, the light of the same standard candle will appear dimmer when it is farther away. GRBs are so powerful that in a few seconds they emit the equivalent of the energy emitted by the Sun during its entire lifetime. Thus, it is possible to observe GRBs at incredibly large distances (a.k.a., high redshift), much further than standard candles like Ia-type supernovae (SNe Ia) that are observed at up to 11 billion light years. Using GRBs as a new type of standard candle will allow astronomers to study and comprehend cosmological issues that could change current models regarding the Universe’s history and its evolution.

Despite decades of observations, a comprehensive model able to explain the underlying physical mechanisms and properties of these objects has not been reached yet. Many possible physical origins for GRBs have been proposed, like the explosion of an extremely massive star (the long duration GRBs) or the merging of two compact objects (the short duration GRBs). Many models about the progenitor responsible of powering GRBs have been proposed as well, such as a black hole, a neutron star (NS) or a rapidly rotating newly born NS with a high magnetific field (magnetar).

Kilonovae (in short: KNe) are astrophysical objects linked to short duration gamma-ray bursts, which are the result of explosions occuring after two very dense objects (for exampe, two neutron stars) merge together. The detection of X-ray emission at a location coincident with the given Kilonovae can also provide the missing observational link between short duration GRBs and gravitational wavesproduced by ssuch stellar mergers. The first detection of the Kilonovae associated with both gravitational waves emission and such a short GRB, namely GRB 170817, has opened a new era of observations and theoretical investigation. The missing piece to this long-standing story is the connection of KNe and the GRB observational correlations that Dainotti et al. now provide.

Figure 2: The LX-T*X-Lpeak relation for the SGRB (short duration GRB) sample with separated KN-SGRB cases. We note here that all the KN-SGRBs (marked in yellow) fall below the best fitting plane. Credit: The Authors.

Even when all the GRBs are observed with the same satellite, in this case the NASA’s Neil Gehrels Swift Observatory, the GRBs’ features are seen to vary very widely over several orders of magnitude. This applies not only to the prompt emission (the main event in the gamma rays), but also to the extended afterglow phase (which follows the prompt emission and is seen over a wide range of wavelengths). Thus, the key point of the article by Dainotti et al., is the hunt for features which remain invariant according to peculiar classes of GRBs.

Figure 3: Histograms of the distance from the Short Duration Plane for KN-SGRBs and SGRBs, considering the correction for selection biases and evolutionary effects. Credit: The Authors.

The team has found a 3-D correlation, i.e. a link between the following three variables that identifies a plane: duration of the X-ray plateau phase, its luminosity, and the luminosity of the peak prompt gamma ray feature. The distances of GRBs from a given class’s plane allowed the authors to determine if GRBs belong to that particular class by showing different features related to this 3-D correlation. The Dainotti et al. study has also shown that although the GRBs-KNe events are a subsample of the larger class of short duration GRBs (red cuboids), they show some observational peculiarities: indeed, they all lie below the short fundamental plane as shown in Figure 2 (yellow truncated icosahedrons). In this analysis, selection biases and evolutionary effects (namely, how the variables change with distance or redshift) have been accounted for, and after correction for selection bias the 3D correlation for GRB-KNe is still tight together with the platinum sample, the tightest sample for the 3D correlation where only well-sampled determined features are taken into account. Thus, both the platinum and the GRBs-KNe plane seems to be the excellent tools for further cosmological studies.

In fact, the GRBs-KNe plane has the smallest observed distance from its plane, called the intrinsic scatter. Here this scatter is 29% smaller than a previous analysis, see Fig. 2, object of a NASA press in 2016, lead by Dr. Dainotti. We note that this finding has been reached in a natural way without assuming any observational criteria, as had been done in Dainotti et al. previous studies. This new result is thus a step much further ahead than previous analyses.

In addition, the separated KNe plane itself still has a very small distance from the 3D plane related to the KNe when evolution is accounted for, see Fig. 3. The smaller the distance is from the plane, the more useful the plane is to be used as a cosmological tool.

A great advantage of using the GRBs associated with Kilonovae is that the GRB-KNe events have a clearer physical emission process compared to other observational GRB classes. Thus, the leap forward in this study is that this sample has a physical grounding related to the fundamental plane relation regardless of the features of the plateau phase which can vary widely from one GRB to another.

References: Prof. Maria Giovanna Dainotti, Aleksander Lenart, Giuseppe Sarracino, Shigehiro Nagataki, Salvatore Capozziello, Nissim Fraija; The X-ray fundamental plane of the Platinum Sample, the Kilonovae and the SNe Ib/c associated with GRBs, ApJ 2020 (DOI: 10.3847/1538-4357/abbe8a). https://arxiv.org/abs/2010.02092

Provided by Astronomical Observatory of the Jagiellonian University

Astronomers Precisely Measure Distance to Magnetar (Astronomy)

XTE J1810−197 (J1810), a magnetar located in the constellation of Sagittarius, was the first magnetar identified to emit radio pulses, and has been extensively studied during a radio-bright phase in 2003–2008. It is estimated to be relatively nearby compared to other Galactic magnetars, and provides a useful prototype for the physics of high magnetic fields, magnetar velocities, and the plausible connection to extragalactic fast radio bursts. Upon the re-brightening of the magnetar at radio wavelengths in late 2018, researchers of current study resumed an astrometric campaign on J1810 with the Very Long Baseline Array, and sampled 14 new positions of J1810 over 1.3 years and has made the direct geometric measurement of the distance to XTE J1810-197, a magnetar located in the constellation of Sagittarius.

An artist’s impression of a magnetar emitting a burst of radiation. Image credit: Sophia Dagnello, NRAO / AUI / NSF.

Magnetars are a variety of neutron stars — the superdense remains of massive stars that exploded as supernovae — with extremely strong magnetic fields.

A typical magnetar magnetic field is a trillion times stronger than the Earth’s magnetic field, making magnetars the most magnetic objects in the Universe.

They can emit strong bursts of X-rays and gamma rays, and recently have become a leading candidate for the sources of fast radio bursts (FRBs).

Researchers performed the phase calibration for the new observations with two phase calibrators that are quasi-colinear on the sky with J1810, enabling substantial improvement of the resultant astrometric precision. They then, combine their new observations with two archival observations from 2006 and have refined the proper motion and reference position of the magnetar and have measured its annual geometric parallax, the first such measurement for a magnetar.

This effect, called parallax, allows astronomers to use geometry to directly calculate the object’s distance. And what they found?

They found that the parallax of 0.40 ± 0.05 mas corresponds to a most probable distance 2.5 (↑+0.4 ↓−0.3) kpc for J1810. Their new astrometric results confirm an unremarkable transverse peculiar velocity of ≈200 km s−¹ km for J1810, which is only at the average level among the pulsar population. The magnetar proper motion vector points back to the central region of a supernova remnant (SNR) at a compatible distance at ≈70 kyr ago, but a direct association is disfavored by the estimated SNR age of ∼3 kyr.

References: H Ding, A T Deller, M E Lower, C Flynn, S Chatterjee, W Brisken, N Hurley-Walker, F Camilo, J Sarkissian, V Gupta, A magnetar parallax, Monthly Notices of the Royal Astronomical Society, , staa2531, https://doi.org/10.1093/mnras/staa2531

Ultra-Diffuse Galaxies Are The Ghosts Of The Cosmos (Astronomy)

When talking about space, people often gravitate to the biggest, the brightest, and the closest to us. But sometimes it’s just as fascinating to learn about the other end: the faint and seemingly insignificant. Ultra-diffuse galaxies are in the latter camp. These dim, wispy galaxies have caught the attention of astronomers, who have sought to find out how they formed.

Ultra diffuse galaxy. Credit: Nasa/ESA


Ultra-diffuse galaxies are fairly large, sometimes stretching to the size of our own Milky Way galaxy. But they contain the same number of stars as your average dwarf galaxy; that is, as little as one one-thousandth the number in the Milky Way. That makes them incredibly dim and hard for astronomers to spot, which is why nearly all of the ultra-diffuse galaxies we’ve discovered so far have been in clusters of other galaxies. The telescopes were already looking at the clusters and spotted those galaxies by happenstance.

But how does a galaxy with so few stars get to be so big? Since the discovery of the first ultra-diffuse galaxy in 2015, astronomers have been split. Some thought that they were ordinary spiral galaxies that contained an unusually large amount of dark matter, while others thought they were just dwarf galaxies that had spread out over a larger area. In November 2016, scientists solved the mystery. By using computer simulations, researchers in Copenhagen and Abu Dhabi were able to watch the formation of nearly 100 virtual galaxies from characteristics they had observed in real ones. What they saw was that young ultra-diffuse galaxies start out like dwarf galaxies but contain a large number of supernovae, whose vast explosions blow other stars and dark matter outward until the whole galaxy is extra large and extra faint. Like a housecat that inexplicably takes up half of a queen-size bed, ultra-diffuse galaxies are dwarf galaxies that have spread out and gotten comfortable.


Dwarf galaxies are the most numerous type of galaxy in the universe. The fact that ultra-diffuse galaxies are dwarf galaxies’ larger sisters means that there are probably many, many more of them out there that we haven’t spotted. Add that to the list of things we never knew we didn’t know.

Supernovae Could Enable The Discovery Of New Muonic Physics (Physics)

A supernova, the explosion of a white-dwarf or massive star, can create as much light as billions of normal stars. This transient astronomical phenomenon can occur at any point after a star has reached its final evolutionary stages.

Supernovae are thought to be associated with extreme physical conditions, far more extreme than those observed during any other known astrophysical phenomenon in the universe, excluding the Big Bang. In supernovae that involve a massive star, the star’s core can collapse into a neutron star, while the rest of it is expelled in the explosion.

During these violent stellar explosions, temperatures in the newborn neutron star can reach over 600 billion degrees, and densities can be up to 10 times greater than those in atomic nuclei. The hot neutron star resulting from this type of supernova is a significant source of neutrinos and could thus be an ideal model for particle physics studies.

For several decades, astronomers and astrophysicists have been trying to prepare for the occurrence of a supernova, devising theoretical and computational models that could aid the current understanding of this fascinating cosmological event. These models could help to analyze and better understand new data collected using state-of-the-art detectors and other instruments, particularly those designed to measure neutrinos and gravitational waves.

Back in 1987, researchers were able to observe neutrinos produced in a supernova for the first and, so far, only time, using instruments known as neutrino detectors. These neutrinos had traveled to Earth over a time period of approximately ten seconds, thus, their observation provided a measurement of the rate at which the remains of a supernova were able to cool down.

For decades now, this measurement was seen as the limit in how quickly exotic particles can cool a supernova remnant. Since it was first introduced in 1987, this point of reference, known as the “supernova cooling constraint,” has been extensively used to investigate extensions of the standard model, the primary theory of particle physics describing fundamental forces in the universe.

Credit: Bollig et al.

Researchers at the Max Planck Institute for Astrophysics in Germany and Stanford University have recently carried out a study investigating the potential of supernovae as platforms to unveil new physics beyond the standard model. Their paper, published in Physical Review Letters, specifically explores the role that muons, particles that resemble electrons but have far larger masses, could play in the cooling of supernova remnants.

The recent study featured in Physical Review Letters was the result of a collaboration between two teams of researchers, one at the Max Planck Institute and one at Stanford. The team at the Max Planck Institute, comprised by Robert Bolling and Hans-Thomas Janka, ran a series of supernova simulations that included Muonic effects, while also incorporating some of the most recent findings about the physics of supernovae.

These simulations led to the creation of the largest existing library of supernova profiles including muons, which is now publicly available and can be accessed by all astrophysics researchers worldwide. Subsequently, De Rocco and the rest of the team at Stanford used this library to compute production rates of axion-like particles, trying to determine where in the parameter space their production would violate the cooling constraint delineated in 1987.

DeRocco, Janka, and their colleagues demonstrated that supernovae could be powerful laboratory models to hunt for new muonic physics, something that was not fully appreciated until now. Their work has already inspired other research teams to seek for exotic physics beyond the standard model by studying muons in supernovae. In the future, this paper could thus pave the way towards new fascinating discoveries about particles in the universe and cosmological phenomena.

References: Robert Bollig et al. Muons in Supernovae: Implications for the Axion-Muon Coupling, Physical Review Letters (2020). DOI: 10.1103/PhysRevLett.125.051104 link: https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.125.051104