Simulating reflected light coronagraphy of Earth-like exoplanets with a large IR/O/UV space telescope: impact and calibration of smooth exozodiacal dust

Observing Earth-like exoplanets orbiting within the habitable zone of Sun-like stars and studying their atmospheres in reflected starlight requires contrasts of $\sim1\mathrm{e}{-10}$ in the visible. At such high contrast, starlight reflected by exozodiacal dust is expected to be a significant source of contamination. Here, we present high-fidelity simulations of coronagraphic observations of a synthetic Solar System located at a distance of 10 pc and observed with a 12 m and an 8 m circumscribed aperture diameter space telescope operating at 500 nm wavelength. We explore different techniques to subtract the exozodi and stellar speckles from the simulated images in the face-on, the 30 deg inclined, and the 60 deg inclined case and quantify the remaining systematic noise as a function of the exozodiacal dust level of the system. We find that in the face-on case, the exozodi can be subtracted down to the photon noise limit for exozodi levels up to $\sim1000$ zodi using a simple toy model for the exozodiacal disk, whereas in the 60 deg inclined case this only works up to $\sim50$ zodi. We also investigate the impact of larger wavefront errors and larger system distance, finding that while the former have no significant impact, the latter has a strong (negative) impact. Ultimately, we derive a penalty factor as a function of the exozodi level and system inclination that should be considered in exoplanet yield studies as a realistic estimate for the excess systematic noise from the exozodi.