Effect of different filtering methods to deal with the granulation noise source
At the end of this year, ESA will launch the CHEOPS satellite, exclusively dedicated to small exoplanets characterisation. The high-precision of this new instrument will allow to refine the planets' radius measurements with a precision inferior to 10%.
However, at this precision level, we expect that the parameters estimation can be biased by the small timescales stellar variability noise sources as the stellar convection. Indeed, convection motions at the stellar surface induce a stochastic colored noise, correlated over timescales of some minutes (granulation) to days (supergranulation), which can reach until half of the transit depth of an Earth transiting a solar analogue. The photometric amplitudes and the typical timescales depends on the stellar spectral type and can be far much higher for F-type exoplanet host stars.
In this contribution, we aim to characterize the properties of granulation at high temporal resolution for FGK stars in terms of intrinsic properties (stationarity), typical timescales of correlation and, resulting amplitudes. To do so, we use currently existing observations of the Sun (21 year of VIRGO observations) and of other stars (Kepler targets observed at short cadence). Based on this understanding, we investigate the use of simple dedicated filtering methods, designed according to the granulation properties, and analyse their performances for planet parameters estimation.