« Visualizing viruses attaching to bacteria: | Main | Highlight: Is Google hurting us? »

Searching for Pulsars

Pulsars are the astronomical equivalents of X-Games athletes.


Pulsar or magical planetoid spouting fairy dust?
You decide.
  1. Have magnetic fields about 5,000,000,000,000 times that of the Earth's.
  2. Have masses larger than the Sun's but diameters the size of a city.
  3. Are consequently as dense as an atomic nucleus (a teaspoon weighs 17 million tons).
  4. Are made mostly of neutrons that were mostly made from electrons and protons being squished together in a stellar collapse.
  5. Can spin up to 716 times a second, sending a pulse in our direction once each spin, like X-Games lighthouses.
Sometimes I stop and try to imagine Washington, DC, spinning faster than a kitchen blender. Really, I try.

Pulsars are sexy science, as science of the extreme variety always is. They have the potential to help us detect gravitational waves, study exotic particle physics from thousands of light-years away, and understand more about the violent, cataclysmic end of the stellar evolutionary process. 

For the past week, and for the next week, I have been and will be helping to teach high school students how to identify new pulsars in survey data taken with the Green Bank Telescope. These students are part of the Pulsar Search Collaboratory Program, a joint venture of NRAO and WVU.

Several years ago, astronomers took data by parking the telescope in one place (well, technically it parked itself, since its track was broken and being repaired) and letting the sky drift over it. With ~130 terabytes of data, they had their hands full enough to share 30 Tb with teenagers.

The survey is processed by PRESTO software, which searches data from each location at which the telescope was aimed (called a "pointing") for periodic signals. The software "folds" the data at many different periods, or adds sections of it together so that signals that repeat with regularity fall on top of each other and add constructively, while noise with no periodicity begins to cancel itself out. If the software (in its infinite, artificial wisdom) believes it sees a repeating signal popping out above the noise, it takes note.

There's a hitch to looking for repeating signals, though: interstellar dispersion. Space does not really live up to its name: it's full of stuff. In this case, we care about the stuff known as "free electrons," which do live up to their name. Radio waves interact with these free electrons, causing them to take longer to get here than they would if there were imprisoned electrons, or no electrons at all. But not all radio waves make it through the sea of negatively charged particles in an equal amount of time: high-frequency waves get through faster than low-frequency waves.
A pulse should line up straight, since it left the pulsar
straight, but you can see in this picture that lower
frequencies (bottom of y-axis) get here later, resulting
in a curved signal.
What does this mean for pulsars? Pulsar are broadband emitters, meaning that they put of radio waves over a range of frequencies, not just at discrete frequencies, like atoms and molecules. So while the pulse all leaves the pulsar at the same exact time, the high-frequency parts get to the telescope before the low-frequency parts, giving astronomers a signal smeared over the time domain, instead of the sharp and peak-like pulse we want to see.

So the magical software can't just search for a bam-bam-bam signal that repeats, because the actual received signal will be more like baaaaaaam...baaaaaaaaam...baaaaaaaam. And that is much more difficult.

The pulses have to be "coherently dedispersed," meaning basically "lined back up." Astronomers must correct for the time delay, and it's only after that correction is performed that pulsars can actually be identified.

The more electrons a radio wave goes through, the more dispersed (spread out over time) it is. But when you're just searching for new pulsars, you don't know how many electrons their photons have been through, so you have to set the software on another task: guess a lot of dispersion measures at the same time that you're guessing a lot of periods, put all those guesses in combination, and see if you can both dedisperse and find something useful to be dedispersing.

When it's done processing (which takes ~24 hours for each pointing), it spits out 30 plots like the one below, contain its 30 best period guesses at various dispersion measures.

This looks like something "the software" could analyze, right? Since it did, in fact, make it?

Well, it turns out that humans' visual pattern recognition is currently better than the bots', especially for pulsars that aren't classic-looking, which are the most interesting ones since they may be sputtering out, speeding up, or orbiting massive companions.

So we bring high school students here for a week and teach them about pulsar properties and how they relate to all the lines and curves and numbers on the plots above.

The first plot is the pulse profile, the result of data folding. The second is also showing the pulse, but spread out over the whole time of the observation (2.5 minutes for each pointing). Right next to that is a reduced chi-squared measurement of how much the data differ from a model of pure noise (hint: hopefully a lot!). The second greyscale image shows whether or not the pulse appears across all the measured frequencies (it should!). The plot below that, which is the last one the students analyze, shows which dispersion measures lined the pulses back up and which didn't.

Students (and students alone) rank these plots and submit interesting ones to astronomers for followup observations.

So far, 5 pulsars have been found by students, at a pulsar-per-gigabyte rate almost identical to that of the astronomers. 

EmailEmail Article to Friend

References (6)

References allow you to track sources for this article, as well as articles that were written in response to this article.

Reader Comments (2)

I wanted to let you know that I had a really great mental image of me with a teaspoon of pulsar and it being so heavy that as soon as I spooned out some pulsar the little metal cup flew to the ground.

I love hearing about things I know nothing about! Great summary of your work!

July 25, 2011 | Unregistered CommenterBrooke N.

The little metal cup flew THROUGH the ground to China, I think.

July 25, 2011 | Unregistered CommenterSarah Scoles

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>