Aside from categorizing galaxies, another component of the Galaxy Zoo project has been asking participants to identify potential supernovae (SNe). The first results are out and have identified “nearly 14,000 supernova candidates from [Palomar Transient Factory, (PTF)] were classified by more than 2,500 individuals within a few hours of data collection.”
Although the Galaxy Zoo project is the first to employ citizens as supernova spotters, the background programs have long been in place but were generating vast amounts of data to be processed. “The Supernova Legacy Survey used the MegaCam instrument on the 3.6m Canada-France-Hawaii Telescope to survey 4 deg2” every few days, in which “each square degree would typically generate ~200 candidates for each night of observation.” Additionallly, “[t]he Sloan Digital Sky Survey-II Supernova Survey used the SDSS 2.5m telescope to survey a larger area of 300 deg2” and “human scanners viewed 3000-5000 objects each night spread over six scanners”.
To ease this burden, the highly successful Galaxy Zoo implemented a Supernova Search in which users would be directed through a decision tree to help them determine what computer algorithms were proposing as transient events. Each image would be viewed and decided on by several participants increasing the likelihood of a correct appraisal. Also, “with a large number of people scanning candidates, more candidates can be examined in a shorter amount of time – and with the global Zooniverse (the parent project of Galaxy Zoo) user base this can be done around the clock, regardless of the local time zone the science team happens to be based in” allowing for “interesting candidates to be followed up on the same night as that of the SNe discovery, of particular interest to quickly evolving SNe or transient sources.”
Remove All Ads on Universe Today
Join our Patreon for as little as $3!
Get the ad-free experience for life
To identify candidates for viewing, images are taken using the 48 inch Samuel Oschin telescope at the
Palomar Observatory. Images are then calibrated to correct instrumental noise and compared automatically to reference images. Those in which an object appears with a change greater than five standard deviations from the general noise are flagged for inspection. While it may seem that this high threshold would eliminate other events, the Supernova Legacy Survey, starting with 200 candidates per night, would only end up identifying ~20 strong candidates. As such, nearly 90% of these computer generated identifications were spurious, likely generated by cosmic rays striking the detector, objects within our own solar system, or other such nuisances and demonstrating the need for human analysis.
Still, the PTF identifies between 300 and 500 candidates each night of operation. When exported to the Galaxy Zoo interface, users are presented with three images: The first is the old, reference image. The second is the recent image, and the third is the difference between the two, with brightness values subtracted pixel for pixel. Stars which didn’t change brightness would be subtracted to nothing, but those with a large change (such as a supernova), would register as a still noticeable star.
Of course, this method is not flawless, which also contributes to the false positives from the computer system that the decision tree helps weed out. The first question (Is there a candidate centered in the crosshairs of the right-hand [subtracted] image?) eliminates misprocessing by the algorithm due to misalignment. The second question (Has the candidate itself subtracted correctly?) serves to drop stars that were so bright, they saturated the CCD, causing odd errors often resulting in a “bullseye” pattern. Third (Is the candidate star-like and approximately circular?), users eliminate cosmic ray strikes which generally only fill one or two pixels or leave long trails (depending on the angle at which they strike the CCD). Lastly, users are asked if “the candidate centered in a circular host galaxy?” This sets aside identifications of variable stars within our own galaxy that are not events in other galaxies as well as supernovae that appear in the outskirts of their host galaxies.
Each of these questions is assigned a number of positive or negative “points” to give an overall score for the identification. The higher the score, the more likely it is to be a true supernova. With the way the structure is set up, “candidates can only end up with a score of -1, 1 or 3 from each classification, with the most promising SN candidates scored 3.” If enough users rank an event with the appropriate score, the event is added to a daily subscription sent out to interested parties.
To confirm the reliability of identifications, the top 20 candidates were followed up spectroscopically with the 4.2m William Herschel Telescope. Of them, 15 were confirmed as SNe, with 1 cataclysmic variable, and 4 remain unknown. When compared to followup observations from the PTF team, the Galaxy Zoo correctly identified 93% of supernova that were confirmed spectroscopically from them. Thus, the identification is strong and this large volume of known events will certainly help astronomers learn more about these events in the future.
If you’d like to join, head over to their website and register. Presently, all supernovae candidates have been processed, but the next observing run is coming up soon!