Absolute Ca II H & K and H-alpha flux measurements of low-mass stars: Extending R '(HK) to M dwarfs

Context. With the recent surge of planetary surveys focusing on detecting Earth-mass planets around M dwarfs, it is becoming more important to understand chromospheric activity in M dwarfs. Stellar chromospheric calcium emission is typically measured using the R '(HK) calibrations of Noyes et al. (1984), which are only valid for 0.44 <= B - V <= 0.82. Measurements of calcium emission for cooler dwarfs B - V >= 0.82 are difficult because of their intrinsic dimness in the blue end of the visible spectrum. Aims. We measure the absolute Ca II H & K and H alpha flux of a sample of 110 HARPS M dwarfs and also extend the calibration of R '(HK) to the M dwarf regime using PHOENIX stellar atmosphere models. Methods. We normalized a template spectrum with a high signal-to-noise ratio that was obtained by coadding multiple spectra of the same star to a PHOENIX stellar atmosphere model to measure the chromospheric Ca II H & K and H alpha flux in physical units. We used three different T-eff calibrations and investigated their effect on Ca II H & K and H alpha activity measurements. We performed conversions of the Mount Wilson S index to R '(HK) as a function of effective temperature for the range 2300 K <= T-eff <= 7200 K. Last, we calculated continuum luminosity. values for Ca II H & K and Ha in the same manner as West & Hawley (2008) for -1.0 <= [Fe/H] <= +1.0 in steps of Delta[Fe/H] = 0.5. Results. We compare different Teff calibrations and find Delta T-eff similar to several 100 K for mid- to late-M dwarfs. Using these different Teff calibrations, we establish a catalog of log R '(HK) and F-H alpha'/F-bol measurements for 110 HARPS M dwarfs. The difference between our results and the calibrations of Noyes et al. (1984) is. log R '(HK) = 0.01 dex for a Sun-like star. Our chi values agree well with those of West & Hawley (2008). We confirm that the lower boundary of chromospheric Ca II H and K activity does not increase toward later-M dwarfs: it either stays constant or decreases, depending on the T-eff calibration used. We also confirm that for H alpha, the lower boundary of chromospheric flux is in absorption for earlier -M dwarfs and fills into the continuum toward later M dwarfs. Conclusions. We confirm that we can effectively measure R '(HK) in M dwarfs using template spectra with a high signal-to-noise ratio. We also conclude that our calibrations are a reliable extension of previous R '(HK) calibrations, and effective temperature calibration is the main source of error in our activity measurements.