-
Overview
Show More Show LessExoplanets, or planets orbiting stars beyond our Sun, can be detected using several techniques. One widely used approach is the transit method, which involves observing brief, periodic decreases in a host star’s brightness as a planet passes in front of it.
Large planets, perhaps similar to Jupiter, which are close to their host stars produce frequent, deep transits which are easy to spot. However, Earth-like planets orbiting Sun-like stars might cause a reduction in the star’s brightness of only about 0.01%, and present a more stringent challenge, particularly since real stars are not perfect light sources but have intrinsic variability due to “starspots” and granulation. Noise and jitter in the detector instruments contribute further difficulties, so that in practice results from multiple transits need to be combined just to be sure the signal is really there.
Traditional approaches to transit detection often involve “phase-folding” the star’s light curve over a grid of trial periods and fitting a geometrically motivated model to the transit profile to estimate the exoplanet’s orbital period. Alternative approaches, such as CETRA, instead search the light curve in time first and phase-fold only the compact results afterwards. A key challenge for all such methods is processing the vast quantities of data generated by large-scale surveys.
The PLATO mission, which aims to discover and study Earth-like planets around Sun-like stars and is scheduled for launch in 2027, will monitor approximately O(105) stars with some observed as frequently as every 25 seconds, over an epoch baseline of at least two years. This presents an unprecedented data-processing challenge.
This Deep Dive presents the CETRA algorithm, devised to perform transit detection at scale for PLATO, and optimised for GPUs. It’s an ideal example because the essential ideas (geometry, time series analysis) are conceptually simple and accessible to non-specialists but clearly highlight the algorithmic choices to be made to unlock the potential of GPUs.
Show More-
Skip to video...
-
Skip to video...
-
Skip to video...
-
Leigh Smith is a Senior Research Associate at the Institute of Astronomy, University of Cambridge, a...
-