Talk detail

Participant	De Bom, Clecio
Institution	Brazilian Center for Physics Research - Dr. Xavier Sigaud, 150 - Rio de Janeiro - Rio de Janeiro - Brazil

Session	OC2	Accepted		Order	Time
Talk	Oral abstract	Title	Finding gravitational arcs in Wide Field Surveys: the Mediatrix Arcfinder and comparison to other methods in a simulated sample
Co-authors
Abstract	Gravitational arcs are powerful probes of the matter distribution in galaxies and galaxy clusters and allow to constrain cosmological parameters. At present time, of order of 102 arcs systems have been discovered in homogeneous samples. This number is expected to increase by an order of magnitude in the next few years due to the upcoming wide fields surveys, such as the Dark Energy Survey (DES), which starting late 2012. Such surveys will cover several thousands square-degrees. Finding arcs in such large areas requires automated algorithms to select arc candidates among hundreds of millions of galaxies and other objects. In this contribution we present a new arcfinding method that uses a neural network to select arc candidates, named MediatrixArcfinder. Objects are identified by SExtractor (Bertin & Arnouts 1996) and have their morphological properties derived from the Mediatrix decomposition (Bom et al. 2012), which are used as input for the neural network. We also present a systematic comparison between this method and three other arcfinders available in the literature: Lenzen et al. (2004), Horesh et al. (2005), More et al. (2012). We determine the efficiency of these arcfinders as a function of arc properties for a sample containing thousands of simulated arcs. We attempt to understand the false positive population using samples from both real data (Canada-France-Hawaii-Telescope Legacy Survey) and simulated images for DES.