nai

Project Progress

Recently, Hsieh & Jewitt (Science, 2006, Volume 312, pp. 561-563) observed three asteroids with comet-like activity inside the main asteroid belt. These Main Belt Comets (MBCs) represent the third known reservoir of comets in the solar system, after the Oort cloud and Kuiper belt, and are a potential source of water for terrestrial planets.

We have been developing techniques and codes to search for MBCs in data produced by Pan-STARRS 1, an all-sky optical survey operated by the University of Hawaii. We anticipate that Pan-STARRS 1 will detect up to 10,000 MBCs, if the 0.3% incidence observed by Hsieh and Jewitt is representative of the Main Belt as a whole.

Because MBCs have a very faint coma, containing only about one percent of the total light of the central point source, Hsieh and Jewitt were forced to use several dozen five-minute stacked images to detect cometary activity. Such image stacking will be impossible to use with a non-targeted survey like Pan-STARRS1. Accordingly, we have developed a method to search for nearly-invisible activity in a single exposure using the brightest polar segment around an MBC. Combining many such detections then allows us to identify MBCs as those asteroids that show a minuscule amount of excess light in a multitude of exposures. This method has proven capable of distinguishing even the faintest MBC 176P from the point sources (stars) in the same field with very high statistical significance.

{{ 1 }}

Figure 1 shows the azimuthal slice scheme used to detect excess light around the brightest of the known MBCs, Elst-Pizarro.

We have been testing our method on the large Thousand Asteroid Light Curve (TALCS) data set of Masiero and Jedicke, obtained at the CFHT telescope.

{{ 2 }}

Figure 2 shows the disagreement of the TALCS asteroids with the null hypothesis of no activity; from the left to right, the panels show the probability of excess light, the probability of non-random directionality (indicative of a consistent tail direction), and the Fisher combination of the two. We do not detect more significant excess light (i.e. large negative log probability) than expected by chance in a sample of 1000 objects, suggesting no obvious MBCs in this limited sample. The known MBCs would have been detected at very high significance.

In additional to this polar slice method, we are working with a graduate student (Sarah Sonnett) on a parallel approach to finding MBCs in the TALCS sample, searching for circular comas visible as a distortion in the point spread function. Both approaches may make it into our final PS1 MBC finding system. This system is expected to begin operations by early 2008.