Observatoire de Paris Institut national de recherche scientifique français Univerité Pierre et Marie Curie Université Paris Diderot - Paris 7

In-flight photometry extraction of PLATO targets : optimal apertures based on noise-to-signal-ratio and frequency of threshold crossing events

mardi 27 novembre 2018, par Victor Marchiori (LESIA)

Mardi 15 janvier 2019 à 11h00 , Lieu : ATTENTION : changement de salle : bâtiment 14

ESA’s PLATO space mission is devoted to unveiling and characterizing new extrasolar planets and their host stars. It will encompass a very large (>2200 deg2) field-of-view, granting it the potential to survey up to one million stars depending on the final observation strategy. Spacecraft telemetry budget cannot handle transmitting individual images for such a huge stellar sample, so the development of an appropriate strategy to perform on-board data reduction is mandatory. We employ mask-based (aperture) photometry to produce stellar lightcurves in-flight. Our aim is thus to find the mask model that optimizes the scientific performance of the reduced data. Three distinct aperture structures are considered. First, we set up a weighted mask delivering global minimum noise-to-signal ratio, computed through a novel fast convergence algorithm. A second weighted mask is built following a Gaussian function. Lastly, we define a mask containing exclusively binary weights. Each strategy is tested on synthetic imagettes generated for 50,000 potential PLATO targets. Stellar population is extracted from Gaia DR2 catalogue. A pioneer criteria is adopted for choosing the optimal solution (structure) : the one providing the best compromise between sensibility to detect true and capability to reject false planet transits, determined based on noise-to-signal ratio and frequency of threshold crossing events. Our results show that, although binary mask statistically presents a few percent higher noise-to-signal ratio compared to weighted masks, all three strategies have nearly the same efficiency in detecting legit planet transits. When it comes to avoid spurious signals from contaminant stars though, binary mask statistically collects significantly less contaminant flux than weighted masks, making the former to be 30% less likely to deliver false transit signatures at 7.1sigma detection threshold. In planet transit finding context thus, choosing apertures based exclusively on how well a transit-like signal can be detected may not be the optimal approach.