Stardust Interstellar Preliminary Examination I: Identification of tracks in aerogel

Abstract Here, we report the identification of 69 tracks in approximately 250 cm 2 of aerogel collectors of the Stardust Interstellar Dust Collector. We identified these tracks through Stardust@home, a distributed internet‐based virtual microscope and search engine, in which > 30,000 amateur scientists collectively performed >9 × 10 7 searches on approximately 10 6 fields of view. Using calibration images, we measured individual detection efficiency, and found that the individual detection efficiency for tracks > 2.5 μm in diameter was >0.6, and was >0.75 for tracks >3 μm in diameter. Because most fields of view were searched >30 times, these results could be combined to yield a theoretical detection efficiency near unity. The initial expectation was that interstellar dust would be captured at very high speed. The actual tracks discovered in the Stardust collector, however, were due to low‐speed impacts, and were morphologically strongly distinct from the calibration images. As a result, the detection efficiency of these tracks was lower than detection efficiency of calibrations presented in training, testing, and ongoing calibration. Nevertheless, as calibration images based on low‐speed impacts were added later in the project, detection efficiencies for low‐speed tracks rose dramatically. We conclude that a massively distributed, calibrated search, with amateur collaborators, is an effective approach to the challenging problem of identification of tracks of hypervelocity projectiles captured in aerogel.