The CARMENES radial velocity (RV) survey is observing 324 M dwarfs to search for any orbiting planets. In this paper, we present the survey sample by publishing one CARMENES spectrum for each M dwarf. These spectra cover the wavelength range 520--1710nm at a resolution of at least $R > 80,000$, and we measure its RV, H$\alpha$ emission, and projected rotation velocity. We present an atlas of high-resolution M-dwarf spectra and compare the spectra to atmospheric models. To quantify the RV precision that can be achieved in low-mass stars over the CARMENES wavelength range, we analyze our empirical information on the RV precision from more than 6500 observations. We compare our high-resolution M-dwarf spectra to atmospheric models where we determine the spectroscopic RV information content, $Q$, and signal-to-noise ratio. We find that for all M-type dwarfs, the highest RV precision can be reached in the wavelength range 700--900nm. Observations at longer wavelengths are equally precise only at the very latest spectral types (M8 and M9). We demonstrate that in this spectroscopic range, the large amount of absorption features compensates for the intrinsic faintness of an M7 star. To reach an RV precision of 1ms$^{-1}$ in very low mass M dwarfs at longer wavelengths likely requires the use of a 10m class telescope. For spectral types M6 and earlier, the combination of a red visual and a near-infrared spectrograph is ideal to search for low-mass planets and to distinguish between planets and stellar variability. At a 4m class telescope, an instrument like CARMENES has the potential to push the RV precision well below the typical jitter level of 3-4ms$^{-1}$.

We report on the first star discovered to host a planet detected by radial velocity (RV) observations obtained within the CARMENES survey for exoplanets around M dwarfs. HD 147379 ($V = 8.9$ mag, $M = 0.58 \pm 0.08$ M$_{\odot}$), a bright M0.0V star at a distance of 10.7 pc, is found to undergo periodic RV variations with a semi-amplitude of $K = 5.1\pm0.4$ m s$^{-1}$ and a period of $P = 86.54\pm0.06$ d. The RV signal is found in our CARMENES data, which were taken between 2016 and 2017, and is supported by HIRES/Keck observations that were obtained since 2000. The RV variations are interpreted as resulting from a planet of minimum mass $m_{\rm p}\sin{i} = 25 \pm 2$ M$_{\oplus}$, 1.5 times the mass of Neptune, with an orbital semi-major axis $a = 0.32$ au and low eccentricity ($e < 0.13$). HD 147379b is orbiting inside the temperate zone around the star, where water could exist in liquid form. The RV time-series and various spectroscopic indicators show additional hints of variations at an approximate period of 21.1d (and its first harmonic), which we attribute to the rotation period of the star.

State-of-the-art Doppler experiments require wavelength calibration with precision at the cm/s level. A low-finesse Fabry-Perot interferometer (FPI) can provide a wavelength comb with a very large bandwidth as required for astronomical experiments, but unavoidable spectral drifts are difficult to control. Instead of actively controlling the FPI cavity, we propose to passively stabilize the interferometer and track the time-dependent cavity length drift externally. A dual-finesse cavity allows drift tracking during observation. The drift of the cavity length is monitored in the high-finesse range relative to an external standard: a single narrow transmission peak is locked to an external cavity diode laser and compared to an atomic frequency. Following standard locking schemes, tracking at sub-mm/s precision can be achieved. This is several orders of magnitude better than currently planned high-precision Doppler experiments. It allows freedom for relaxed designs rendering this approach particularly interesting for upcoming Doppler experiments. We also show that the large number of interference modes used in an FPI allows us to unambiguously identify the interference mode of each FPI transmission peak defining its absolute wavelength solution. The accuracy reached in each resonance with the laser concept is then defined by the cavity length that is determined from the one locked peak and by the group velocity dispersion. The latter can vary by several 100m/s over the relevant frequency range and severely limits the accuracy of individual peak locations. A potential way to determine the absolute peak positions is to externally measure the frequency of each individual peak with a laser frequency comb (LFC). Thus, the concept of laser-locked FPIs may be useful for applying the absolute accuracy of an LFC to astronomical spectrographs without the need for an LFC at the observatory.