CARMENES Exoplanet Survey

Updated 25 July 2025

CARMENES Exoplanet Survey is a comprehensive radial-velocity program using dual-channel spectrographs to detect terrestrial and low-mass exoplanets around nearby, bright M dwarf stars.
It employs ultra-stable, fiber-fed spectrographs with m/s precision and automated pipelines to mitigate stellar activity and instrument drifts for robust planet detection.
The survey’s extensive, homogeneously processed dataset enables statistical studies of planet occurrence, atmospheric characterization, and insights into planet formation around low-mass stars.

The CARMENES Exoplanet Survey is a comprehensive radial-velocity program and associated instrumentation initiative targeting the detection and characterization of planets—particularly terrestrial and low-mass planets—orbiting nearby, bright M-dwarf stars. Built on the foundation of an advanced, dual-channel spectrograph and a multinational scientific collaboration, CARMENES has supplied an unprecedented dataset for statistical, theoretical, and atmospheric studies of exoplanets in the solar neighborhood, with a special focus on the environments most conducive to habitability.

1. Instrumentation and Technical Design

CARMENES comprises two high-resolution, fiber-fed echelle spectrographs mounted on the 3.5-m telescope at Calar Alto Observatory in Spain. The instrument covers a continuous wavelength range from 0.52 to 1.71 μm, split into visible (VIS: 0.52–0.96 μm, R ≃ 95,000) and near-infrared (NIR: 0.96–1.71 μm, R ≃ 80,000) channels (Caballero et al., 7 Mar 2025, 1210.5465). Both spectrographs are installed in ultra-stabilized, vacuum-enclosed, temperature-controlled tanks (VIS at near ambient, NIR at ~–135°C), crucial for achieving RV precision of 1 m/s or better. Extreme temperature stability is required: for the NIR spectrograph, a stringent ±0.01 K stability at a working temperature of 243 K minimizes instrumental drifts to meet m/s-level RV precision (1007.3682).

Calibration systems for both channels include hollow-cathode lamps, tungsten flat-field lamps, and Fabry–Pérot etalons. The dual-channel design is especially notable: it maximizes photon throughput for M dwarfs—whose emission peaks from VIS to NIR—and enables simultaneous collection of broad spectral data, fostering both high-precision RV extraction and robust monitoring of activity indicators.

2. Survey Strategy and Methodology

CARMENES targets a statistical sample of several hundred nearby (within ~25 pc), bright (typically J < 10 mag) M dwarfs (1210.5465, Reiners et al., 2017). Target selection is deliberately broad in sky coverage and avoids significant pre-screening for activity or age (except for binarity and extreme youth), yielding a representative sample for occurrence rate and population studies (Ribas et al., 2023).

Survey observations are scheduled using a multi-tier, evolutionary computation scheduling tool called CAST, which maximizes the fraction of telescope time spent acquiring data, enforces uniform coverage, and optimizes cadence to enhance detectability of planetary-period signals in the presence of stellar noise (Garcia-Piquer et al., 2017). Data acquisition typically yields dozens to over 100 high-resolution spectra per target.

Data processing is highly automated, leveraging the caracal pipeline for basic reduction and specialized pipelines (e.g., SERVAL) for RV measurement and activity index extraction. Advanced RV extraction exploits template-matching or cross-correlation algorithms and applies nightly zero-point corrections to combat instrumental drifts over multi-year baselines. Careful mitigation of telluric absorption is implemented using spectral masks (Ribas et al., 2023).

3. Achievements in Planet Discovery and Population Statistics

The survey has confirmed or discovered dozens of planets, ranging from Earth-mass up to giant planets, including:

Super-Earth and habitable-zone candidates (e.g., Teegarden’s Star, HN Lib b) (González-Álvarez et al., 2023).
Well-characterized multi-planet transiting systems (e.g., LTT 3780 b/c) (Nowak et al., 2020).
Detection and refinement of parameters for gas giants in wide orbits around M dwarfs (e.g., GJ 3512 b/c, GJ 9773, GJ 508.2) (Morales et al., 21 Jul 2025).
Rare discoveries of giant planets orbiting very low-mass and/or metal-poor hosts, challenging the predictions of classical core accretion theory (Quirrenbach et al., 2022, Morales et al., 21 Jul 2025).

Statistical analysis shows an average of $\overline{n}_{\rm pl} = 1.44 \pm 0.20$ planets per star (in the range $1\,M_\oplus < M_p\sin i < 1000\,M_\oplus$ , $1\,\text{d} < P_\text{orb} < 1000\,\text{d}$ ), indicating that almost every M dwarf hosts at least one planet (Ribas et al., 2023). Occurrence rates for giant planets with periods up to two years are estimated at 2–6% (Quirrenbach et al., 2022). The survey also provides robust occurrence rates for planets in and near the habitable zones of low-mass stars.

4. Methods for Stellar Activity Handling and Planet Validation

Stellar magnetic activity—leading to line shape variations and RV “jitter”—presents a primary astrophysical noise source for M-dwarf surveys (Schöfer et al., 2019, Jeffers et al., 2022). CARMENES addresses this by:

Simultaneously recording VIS and NIR spectra, exploiting different activity sensitivities with wavelength;
Extracting a suite of chromospheric and photospheric activity indicators, including Hα, Ca II infrared triplet, Na I D lines, He I, TiO/VO bands, the chromatic index (CRX), and differential line width (dLW);
Applying advanced time-series modeling (e.g., Gaussian process regression with kernels such as the simple harmonic oscillator or quasi-periodic form) to jointly fit planetary and activity-driven signals (Jeffers et al., 2022, Amado et al., 2021);
Utilizing low-resolution Doppler imaging to reconstruct the starspot distribution and quantify the effect of surface inhomogeneities on the measured RV (Jeffers et al., 2022);
Incorporating multi-wavelength and multi-instrument RVs as well as photometric time-series to cross-validate planetary hypotheses and disambiguate rotation-induced and Keplerian signals.

This comprehensive activity analysis is critical for robust planet confirmation, particularly for low-mass planets with RV amplitudes near the stellar activity “floor.”

5. Advancements in Stellar and Planetary Characterization

Beyond planet detection, CARMENES data support detailed atmospheric and physical characterization:

High-resolution, multi-band spectral atlas for >300 M dwarfs, enabling precise determination of effective temperature, surface gravity, metallicity, projected rotational velocity, and atmospheric parameters of the host stars (Passegger et al., 2016, Reiners et al., 2017).
Compilation of a public, multi-band photometric and SED-based catalog of nearby dwarfs, with homogeneous estimates of luminosities, radii, and masses; critical for deriving precise planetary properties from transit and RV data (Cifuentes et al., 2020).
Direct empirical calibration of the mass–radius–luminosity relations in the low-mass regime, including the discovery and characterization of new double-line spectroscopic binaries (Baroch et al., 2018).
Transmission and emission spectroscopy of exoplanet atmospheres in the NIR, including detection of atomic helium, with implications for atmospheric escape and composition.

6. Results on Planet Formation and Demographics

CARMENES results contribute significantly to the understanding of planet formation around low-mass stars:

Giant planets are rare in close orbits but appear to be more common than predicted in wide orbits around metal-poor M dwarfs—suggesting formation or migration mechanisms (such as slower migration or disk fragmentation) distinct from those invoked for FGK stars (Morales et al., 21 Jul 2025, Quirrenbach et al., 2022).
The survey documents the occurrence of compact, multi-terrestrial planet systems and a substantial population of sub-Neptunes and super-Earths populating the “radius valley,” consistent with atmospheric loss models and core accretion formation (Nowak et al., 2020, Amado et al., 2021).
Planet discoveries in the habitable zones (e.g., HN Lib b, HD 180617 b) (González-Álvarez et al., 2023, Kaminski et al., 2018) enable tests of planet size, density, and composition across divergent irradiation and metallicity regimes.

7. Consortium Structure and Legacy

CARMENES represents a major German–Spanish instrumentation and science partnership (Caballero et al., 7 Mar 2025, 1210.5465). The bilateral consortium coordinated all aspects of instrument development, science planning, and survey execution. The project’s multi-year guaranteed time led to the release of large, homogeneously processed data products (calibrated spectra, RVs, and activity indicators) in public data releases (Ribas et al., 2023). The instrument’s continuous upgrades and ongoing surveys ensure its continued relevance in coordinated ground-, space-, and astrometric (e.g., Gaia) investigations.

The CARMENES Exoplanet Survey is thus a cornerstone facility combining precision radial-velocity searches, spectroscopic and activity diagnostics, and survey design, contributing fundamentally to both the observational and theoretical frameworks of planetary science around the Galaxy’s most common stars.