NASA Astrobiology Institute (NAI)

We developed new image stacking techniques to attempt to recover comet 85P/Boethin as a target for the NASA EPOXI mission (retargeting of the Deep Impact probe). We were unable to recover Boethin using many nights of observations from the CFHT, Subaru, VLT, and Gemini telescopes. We found one candidate, but it was not deemed certain enough to justify retargeting EPOXI. At the time of this writing, the object has not been recovered, suggesting it may have broken up.

The Deep Impact Extended Mission was the planned retargeting of the Deep Impact probe to visit comet 85P/Boethin. A crucial first step to the mission was the recovery of 85P, in order to make the necessary course corrections for the flyby. We have been working on image data processing and analysis for the Deep Impact recovery.

We obtained extensive imaging along 85P’s line of variations (LOV) using the half-degree field Suprime-Cam CCD array on the Subaru telescope, the one degree CFHT Megacam camera, and smaller fields using Gemini GMOS and the ESO VLT. Combining images on this large scale is complex, necessitating the removal of significant optical distortions, the computation of precise global coordinates, and the combination of the chip array into a global image, including static sky subtraction and star-masking. We have written specialized software to accomplish these tasks.

Because the LOV covers a large fraction of a degree on the sky, the location of the comet is poorly constrained. More crucially for the image stacking needed to detect a faint object, the comet’s point in its orbit, and thus its velocity, are poorly constrained as well. Thus it is impossible to stack the images at a fixed rate, which is valid only for one possible position on the LOV. Accordingly, we have developed a novel stacking approach that effectively distorts each image in a manner that maps the LOVs from different images onto each other using a global coordinate transformation, ensuring that the comet is detected regardless of its position along the LOV. We have also developed a novel “DNA banding” method to translate and juxtapose the LOV from different telescopes, to see if a band consistent with several low-significance detections of the comet is visible.

Figure 1 shows our low-significance stacked CFHT detection of a candidate, with synthetic objects for stacking verification; and Figure 2 shows a DNA band map from several telescopes.

This image processing project failed to find 85P/Boethin, although its validity was demonstrated by detecting other objects moving at close to Boethin’s rate. Boethin has not been recovered near its expected location this year, supporting our tentative conclusion that it has broken up.