In light of recent events at my favorite observatory, a thread on one of my favorite #AODiscoveries: pulsar swooshes! #AreciboLove

Arecibo is a favorite of many astronomers because it's one of the most sensitive radio telescopes on the planet! One day, Dr. Joanna Rankin was using it to observe two radio pulsars, B0919+06 and B1859+07, and noticed something weird...

Normally, the emission of pulsars comes from the same longitude, or the same spot on the star, but these pulsars didn't seem to be following that.

She and colleagues discovered that occasionally the light would come earlier (which corresponded it seeming to come from a different longitude) for a few pulses, then transition back.

This figure (figure 1 from Rankin et al. 2006, https://arxiv.org/pdf/astro-ph/0601368v1.pdf) shows these "shifts." X-axis: longitude, y-axis: pulse number, colors: different intensities/values. The different columns show the different Stokes parameters/polarizations (total intensity on the far left).

Pulsar B0919+06 is on the left and you can see one "event" near the top, but you can see multiple shifts B1859+07 on the right. Nothing like this had ever been seen in any pulsar.

I love studying pulsars because they do weird things. They can null (basically turn off for a period of time) and mode (where they change emission patterns). This phenomenon was especially weird though.

Rankin and co. tried to explain these using mode-changing or other emission physics, but couldn't pinpoint an ultimate cause and published their paper (for more information, paper here: https://arxiv.org/pdf/astro-ph/0601368v1.pdf and also feel free to comment here and I'll explain).

Enter: a very enthusiastic undergrad. When presented with this project, Dr. Rankin (my advisor) said these might be caused by asteroid belts. I picked the project basically because I could explain it easily to my friends. I had to idea about the excitement I was in for....

My first job was to characterize these events, which were informally called "swooshes"!

During these swooshes:
1) the regular pulsed emission gradually moved earlier in longitude over a few pulses
2) remained early for a dozen or more pulses—largely emptying the usual emission window, and
3) then returned over a few pulses to the usual longitude.

They affected all of the light of the pulsar, not just a little bit. They also looked the same in both pulsars, so we knew the same phenomenon was causing them.

Below, close-ups of these pulsars (left: B0919+06, right: B1959+07) (figures 2 and 3 from Wahl et al. 2016, https://arxiv.org/pdf/1607.01737.pdf)

One of the weirdest things about them is that in B1859+07, they come in lots of different shapes, particularly in B1859+07 (name credits: Han et al. 2016 https://arxiv.org/pdf/1601.02889.pdf and Wahl et al. 2016)!

I tried to look for patterns in the types, seeing if one type occurred every X number of swooshes, but found nothing. These things were so weird and confusing!!

So one day I was sitting at my desk and one of Dr. Rankin's previous students looked at the frequent swooshes in B1859+07 and went, "Let's try this....."

In these swooshes we didn't think had any pattern to them, suddenly this really sharp periodicity arose out of them....the sharp peak corresponds to a periodicity of about 153 pulses....(part of figure 4 in Wahl et al. 2016, https://arxiv.org/pdf/1607.01737.pdf)

In addition to being really excited, we were also really confused. So these didn't look periodic but apparently had a strong periodicity when a periodogram? Weird! We also checked B0919+06 and the period was ~700 pulses but it was much rougher (e.g some would appear every ~1400).

So we started thinking.....what if the periodicity of these swooshes corresponded to the period of an orbit....?

Using Kepler's 3rd law (below), where M is the total mass of the star and companion, P is the period of the swooshes, and "a" is the semi-major axis of the orbiting body, we worked out what the distance from each star to each companion would be given the swoosh periodicity....

If this was an orbiting body, B1859+07’s companion would be ~27,000 km from the pulsar's surface. That's crazy, because the light cylinder (the radius at which something co-orbits the star at the speed of light), is 21,000 km, so it's going around REALLY fast.

The companion, if it existed, would be at ~120,000 km (light cylinder = 31,000 km), so a bit further (which meant it went around more slowly) but that's still REALLY close....

According to a paper by Ryan Shannon et al. (link here: https://arxiv.org/pdf/1301.6429v1.pdf), asteroid belts can only survive at distances greater than 1 AU (150,000,000 km)....so that means these can't be asteroid belts!

In order for the companion to stay together and not break up due to the enormous amount of gravity of a pulsar, it would have to have a density of ~10^5 g/cm^3....which is the density of a white dwarf core.

In short, if the swooshes were caused by an orbiting companion, they would go around the stars super fast (once every ~99 seconds for B1859+07 and ~300 seconds/5 minutes for B0919+06), faster than any known binary. However.....

These periodicities are really rough. They don't happen exactly every ~153 pulses and ~700 pulses, the periodicities are really weird (figure 4 of Wahl et al. 2016, https://arxiv.org/pdf/1607.01737.pdf).....

There are so many questions here! The periods are incredibly short, how did these form? Will they evolve? Also, why does the emission move earlier in the pulsars? Why not later? So many questions here.

The idea that swooshes are caused by white dwarf cores orbiting around these pulsars is just a hypothesis to explain these mysterious events!

Probably THE most exciting part of this project (/of my undergrad career): sitting in my office one afternoon and getting an email from a writer at @newscientist saying he wanted to interview us about our paper. Article here: https://www.newscientist.com/article/2097678-mysterious-swoosh-caused-by-pulsars-hugging-companions-close/