Speaker
Description
Stellar activity is today the main limitation to detecting Earth-like planets orbiting Sun-like stars with the RV method. The effects of plages and spots strongly modified high-resolution spectra introducing RV signals difficult to model. Part of the problem is related to the absence of good activity proxies for the photospheric faculae and spots active regions.
By using the CaII H&K lines of three different solar instruments, we showed that a unique profile is expected from each component in the chromosphere. Both emission profiles are strongly collinear, but not identical, giving some possibility to disentangle the contribution of plages/spots.
This conclusion was then extrapolated for Alpha Cen B, a K-dwarf star, by using a machine learning dimensionality reduction algorithm (independent component analysis - ICA) on the spectra time series. The ICA reveals that the time domain vectors recovered (namely the plage and spots filling factor modulation) are new activity proxies powerful to correct the RVs, contrary to the classical Mont Wilson S-index.
Submit to 'solar physics' topical issue? | Maybe |
---|