- FRBs last only a few thousandths of a second at most;
- large radio telescopes typically see only a small fraction of the sky at one time;
- they emit only at radio wavelengths, with apparently no counterpart at optical wavelengths;
- the majority of them only appear once, and the few that do repeat have no obvious periodicity to them.
Complicating matters further, the pulse from an FRB arrives at a later times at lower frequencies due to a phenomenon known as dispersion; the more ionised particles there are between us and the FRB, the larger this effect becomes. We don't know in advance how much dispersion a non-repeating FRB will experience, so detecting one in "real-time" through a continual sweep of likely dispersion ranges requires serious computing power. Thanks to telescopes like Parkes, UTMOST, ASKAP and CHIME we now know of hundreds of FRBs.
One of the other great mysteries in astronomy has been the "missing baryon problem". Observations of the cosmic microwave background left over from the Big Bang allow cosmologists to get a fairly good estimate of the fraction of the Universe made up of baryons, i.e. everything other than dark energy and dark matter. But when they conducted a "census" of all the baryons that they could detect, they found barely half of the expected 5%. They hypothesised that the "missing" baryons most likely exist in a very hot, diffuse and ionised state between galaxies, making them next to impossible to detect in traditional ways.
It wasn't long after the discovery of the first FRBs that scientists including Jean-Pierre Macquart realised that the very property that makes FRBs so hard to detect, namely the amount of dispersion induced by their passage through any ionised material, would make FRBs an ideal tracer of these missing baryons. But an FRB detection by itself tells us only how many baryons in total the burst passed through - to convert this to an actual density, one needs to know how far the burst has traveled. And for that you'd need to firstly identify the galaxy it originated from, and then measure its distance via its redshift and the Hubble constant. But until 2 years ago hardly any FRBs had been traced back to their host galaxies. This all changed on 24 Sep 2018, when the Australian-led Commensal Real-time ASKAP Fast Transients (CRAFT) survey team successfully used the ASKAP array in Western Australia to pin down the location of an FRB from just a single pulse to an accuracy of less than 1 arcsecond.
Macquart and the CRAFT team used mainly the ESO VLT, as well as the Keck, Gemini, and Magellan telescopes to measure the redshifts for the host galaxies of the first half-dozen FRBs localised with ASKAP. When they plotted the FRB total dispersion against redshift they got this:
The other thing you might conclude from this plot is that FRBs make lousy distance indicators, by comparison with (say) Type Ia supernovae (whose progenitors incidentally we don't fully understand either). Much of that scatter though is not due to measurement error - we can measure dispersions and redshifts to better than 0.1%. That scatter indicates that baryons are not uniformly distributed in space, but are more likely to be in filaments and shells surrounding voids. Once we have similar measurements for thousands of FRBs in all different directions we'll be able to trace out this structure and enter a whole new realm of "cosmic tomography", from which we could learn not only how gas flows into and out of galaxies, but also start to apply FRBs as "cosmic rulers" at redshifts beyond 2. Australia is leading the way through a combination of ASKAP and ESO, and the dedicated efforts of the international CRAFT team.