Stellar activity index Ca II H&K as RV noise predictor

Answer

The only explicit numeric cut in this corpus is log R'HK of about -4.7: in the HARPS FGK planet-search sample, active stars with log R'HK ≳ -4.7 that could not be eliminated from CORALIE measurements were screened early and only rarely measured b08014b5. It sits beside two other criteria, v sin i < 3-4 km/s and exclusion of binaries or multiples, applied to minimize RV jitter b08014b5. No S-index threshold is stated in the corpus.

Summary

Quiet stars in the HARPS sample cluster near log R'HK = -5.0, and the Sun at cycle minimum is about -4.97 b08014b5. The cut at -4.7 is therefore a screen for clearly active stars and not a fine quietness criterion. The second fact, on the HARPS-N Rocky Planet Search, shows the opposite practice: activity is classified qualitatively from long Mount Wilson S-index records, with no numeric cutoff given 05d4a201. The two facts do not conflict; they describe different surveys and the corpus does not support a single universal threshold.

Evidence

• Lovis et al. 2011 (HARPS, about 400 stars) — activity screen at log R'HK ≳ -4.7, plus v sin i < 3-4 km/s and no binaries; G dwarfs accumulate near -5.0, K dwarfs spread between -4.7 and -5.0, and no separate cut per spectral type is stated. b08014b5
• Anna John et al. 2023 (HARPS-N Rocky Planet Search) — HD 166620 is in a Maunder-minimum state (S-index roughly 0.18-0.20 over a 57-year record from 1966) and HD 144579 is moderately active; detection of Msin(i) <= 1 Earth mass at periods under 10 days was reached for the quieter target. The retrievable text states no numeric log R'HK or S-index selection threshold. 05d4a201

Methodology

The two facts were kept separate because they come from different surveys and different instruments, and only one reports a numeric screen. The second was read as a negative finding about what that paper states, not as evidence that no such threshold exists for the programme.

Limitations

• Only two facts; the -4.7 figure rests on one survey paper.
• Neither fact ties a threshold to Earth-mass detectability quantitatively; -4.7 is a screening level in a sample design, not a derived limit.
• No S-index (as opposed to log R'HK) cutoff appears anywhere in the corpus.
• Other surveys' cuts (for example other planet-search programmes) are not covered.