Title: Four New Fast Radio Bursts Discovered in the Parkes 70-cm Pulsar Survey Archive
Authors: F. Crawford, S. Hisano, M. Golden, T. Kikunaga, A. Laity, D. Zoeller
First Author’s Institution: Franklin & Marshall College
Status: Accepted for publication in MNRAS
Fast radio bursts (FRBs) have been popping up all over the town the last few years. These enigmatic events are short (a few milliseconds), super energetic (~10x the luminosity of the Sun) bursts of radio emission originating from other galaxies (although this Astrobite covers an FRB-like burst from within our own galaxy). While the first FRB was discovered in 2007, over 3000 of these events have now been detected. However, despite the high numbers of FRBs that we’ve found, we still don’t know a lot about them. For example, we have a few ideas for what might create these super powerful events (e.g. maybe a highly magnetized neutron star spinning super fast), but we still haven’t settled once and for all on a theory for their origins and the physical mechanisms causing the bursts.
Looking into the past
In an attempt to unlock the mystery behind these events, today’s authors take a look at radio survey data from the 1990s (we’re bringing it way back today) to try to find more FRBs. The survey goes by the name ‘Parkes 70-cm pulsar survey’ and, as the name suggests, it was a radio survey at the Parkes Observatory looking for pulsars (rapidly rotating neutron stars) at a frequency of 436 MHz. This survey was widely successful, finding 101 new pulsars! However, when looking for radio transients such as pulsars, you need to consider something known as the dispersion measure of the pulsar. In fancy terms, the dispersion measure (abbreviated DM) of a given source is the integrated electron density between you and the source. In non-fancy terms, it is an approximation for distance: a larger DM typically means farther away. Why does this matter? Well, when radio waves propagate through the interstellar medium, they are slowed down by electrons. Lower frequencies are slowed down more than higher frequencies, and thus there is a delay between the arrival times of low and high frequency emission. The DM is used to quantify this delay. When searching pulses from a given source, you need to re-align to the pulse (using the DM) to get the maximal signal to noise (in order to make an actual detection).
But what do you do when you’re looking for new sources with unknown DMs? Things get computationally expensive! When you’re looking for unknown sources, you have to search over a wide range of DM values and re-align the pulse multiple times until you maximize the signal to noise. Since the ‘Parkes 70-cm pulsar survey’ was only looking for pulsars, they only searched over the DM range expected for our Galaxy. They didn’t consider sources that had gone through the electron density content of multiple galaxies or the intergalactic medium. But, FRBs are extragalactic events, meaning you need to search for DMs much larger than that of our galaxy to find them. This is where today’s author’s re-analysis comes in.
The authors both search over a DM range that is ~5X as large as that from the ‘Parkes 70-cm pulsar survey,’ and search for temporally wider FRBs, up to 150 ms. The span in time and space has again been limited due to computational costs — if only we had unlimited computational resources! They use two different search algorithms to look for possible FRBs in their data, and come up with a whopping ~75,000 candidates. They then compare these candidates against the known-pulsar database, and manually check by eye any which aren’t from known pulsars. Of the 75,000, they can only conclusively classify four as FRBs (can you say huge data reduction).
Four out of 3000, how special can they be?
If you remember how many FRBs I said we had found in the beginning (~3000), you might be thinking that four seems relatively small compared to the total number discovered. And you’re right! But these four FRBs are fairly special in three ways:
1. Of current known FRBs, they have the oldest discovery date (they were found in data from the 1990s).
2. One has the highest DM discovered yet (3338 pc cm^-3 for the FRB/pulsar folks out there)
3. They have some of the longest durations (201 ms, 157 ms, 113 ms, and 51 ms) discovered yet (see the figure below).
The authors mainly focus on this third point, and in particular how much of the width of the pulse is intrinsic to the source vs. 1. arising from interstellar scattering (scattering of radio waves due to free electrons in the interstellar medium) or 2. intra-channel dispersion smearing (broadening in time due to limited frequency resolution of a telescope). They find that the contributions from both interstellar scattering and intra-channel dispersion smearing would be close to negligible, and thus the long durations of the bursts are likely intrinsic, not extrinsic.
What about shorter duration FRBs? And what does all this mean?
You might be a bit suspicious that all the FRBs they detect are super long in duration as the so far known sample of FRBs is much shorter in duration. The authors discuss this briefly, and the most plausible explanation is that the intra-channel dispersion smearing (which increases with DM) estimated for this survey is significant. FRBs would not be detectable if their width was smaller than the contribution from intra-channel dispersion smearing, which in this case is ~7 ms for an FRB with a typical DM of ~500 pc cm^-3.
So, the authors found four FRBs with particularly long durations. What does this mean for the rest of the FRB population? They suggest that there might be a large population of long (comparatively speaking) FRBs just waiting to be discovered. And, there might be even more FRBs lurking in other past radio surveys. So, let’s get looking!
Astrobite edited by Olivia Cooper
Featured image credit: Daniel John Reardon/CC BY-SA 4.0
Hi, very interesting article.
I’m a bit confused about their explanation. At first, they say that “dispersion smearing” is negligible, so the long duration of the bursts is an intrinsic characteristic, but then they explain the phenomenon as “significant dispersion smearing”?