Taking an image of a planet proves to be extremely challenging when the planet orbits a star tens of parsecs away from our sun. Light signals from an Earth-like exoplanets are 10 orders of magnitudes dimmer than signals from their host stars and can only be detected by space telescopes employing a special optical design – the coronagraph. However, the fraction of the starlight blocked by coronagraphs suffers from manufacturing imperfections, alignment errors, and mechanical and thermal stresses acting on the instrument. These issues have to be addressed in real-time via estimation and control of the wavefront of the incoming light, as well as with offline image post-processing techniques.
The choice of observation and post-processing strategies has significant effect on the planet detectability threshold. In particular, statistically informed reduced-order approaches may reduce the severe wavefront stability requirements on future space telescopes, making them more affordable and facilitating the discovery and characterization of habitable exoplanets.
