Würzburg 2018 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

EP: Fachverband Extraterrestrische Physik

EP 8: Poster Session

EP 8.4: Poster

Wednesday, March 21, 2018, 17:00–18:30, BSZ - Pabel HS

Error analysis of telescope observations using Monte Carlo tests — •Stephan Schlegel and Joachim Saur — Universität zu Köln

Telescope observations of faint objects are afflicted with multiple errors, both statistical and systematical. To investigate structures on those objects, their significance has to be calculated. Since the error sources are usually correlated, the analytical forward propagation is often difficult and nontransparent. To address these problems, a Monte Carlo test for a synthetic dataset containing a superposition of two Poisson distributed signals, one for each, background and target signal, was carried out. The background field is considered to be either homogeneous or with a gradient. Furthermore, systematical errors of the detector were added. The aim of this approach is to estimate the significance of the target signal observed by the Hubble Space Telescope using processed telescope data for the Monte Carlo test algorithm including the internal error sources of the instrument. Here we focus on the comparison between commonly used analytical calculation of signal to noise ratio, which were biased by simplifications and assumptions with the newly developed Monte Carlo approach.