Press Release
Open Access
Issue
A&A
Volume 670, February 2023
Article Number A139
Number of page(s) 23
Section Catalogs and data
DOI https://doi.org/10.1051/0004-6361/202244879
Published online 22 February 2023

© The Authors 2023

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication.

1 Introduction

M-type dwarfs provide some advantages with respect to Sunlike stars in the search for exoplanets, particularly those with low masses. Their relatively small sizes and masses result in stronger planetary signals. Furthermore, their low intrinsic luminosities imply that temperate planets orbiting within their liquid-water habitable zone have shorter orbital periods, of the order of tens of days (Kopparapu et al. 2013). In addition, they constitute an abundant stellar population, comprising the majority of stars (78.5%) in the solar neighbourhood (Reylé et al. 2021). The main drawbacks of M dwarfs as targets for exoplanet searches are their intrinsic faintness and the fact that a relatively large fraction of them show magnetic activity phenomena, especially the later spectral types (Reiners et al. 2012). A number of efforts have successfully exploited the so-called M-dwarf opportunity for planet detection over the past few decades (e.g. Delfosse et al. 1999; Endl et al. 2003; Wright et al. 2004; Bonfils et al. 2005; Nutzman & Charbonneau 2008; Zechmeister et al. 2009; Johnson et al. 2010; Ricker et al. 2015; Affer et al. 2016; Seifahrt et al. 2018; Bayliss et al. 2018).

The CARMENES1 instrument and survey were specifically conceived to search for temperate rocky planets around a sample of nearby cool stars (Quirrenbach et al. 2014). The spectrograph was designed to provide high-accuracy radial velocity (RV) measurements with long-term stability in a broad wavelength interval where M-dwarf stars have the peak of their spectral energy distribution. Moreover, such wide coverage provides a range of stellar activity indicators to assess the nature of potential RV signals as well as valuable spectral information that can be used to characterise the stellar targets.

The CARMENES instrument is installed at the 3.5 m telescope of the Calar Alto Observatory in Almería, Spain (37°13′25″N, 2o32′46″W). It provides nearly continuous wavelength coverage from 520 nm to 1710 nm from its two channels: the visual channel (VIS), with a spectral resolution of R = 94600, covers the range λ = 520–960 nm, while the near-infrared channel (NIR) yields a resolution of R = 80 400 within a wavelength interval λ = 960–1710 nm (Quirrenbach et al. 2016). Both channels are coupled to the telescope by optical fibres, with a projection of 175 on the sky.

A sample of about 350 M dwarfs across all M spectral subtypes comprises the targets of the main survey. A total of 750 useful nights were reserved as guaranteed time observations (GTO) for the CARMENES consortium, and these ran for 5 yr, from 1 January 2016 to 31 December 2020.

The present publication accompanies the release of the observations acquired with the CARMENES VIS channel over the course of the RV survey within the GTO programme, which we have dubbed the CARMENES Data Release 1 (DR1). This includes raw data, calibrated spectra, and high-level data products, such as RVs and spectroscopic indicators. The paper is structured as follows. Section 2 describes the design and execution of the CARMENES survey. In Sect. 3, we present the CARMENES GTO target sample and provide a description of its statistical distribution. Section 4 describes the observations collected within the GTO and the processing data flow from raw frames to calibrated RVs and ancillary data products. In Sect. 5, we discuss the properties of the CARMENES DR1 regarding internal and external precision, we provide information regarding the presence of periodic signals in the data, and we present and discuss the sample of exoplanets in the surveyed targets. Furthermore, we present revised planet occurrence rates considering all publicly released data. Finally, Sect. 6 provides the summary and conclusions of the work.

2 The CARMENES survey

The initial goal of the GTO survey was to collect approximately 70 spectra for each of the foreseen 300 targets (Garcia-Piquer et al. 2017), which would have yielded a grand total of ~21 000 spectra. During the survey, we identified a number of targets with high-amplitude RV variations (RV scatter >10 m s−1 and υ sin i > 2 km s−1), which we classified as RV-loud (Tal-Or et al. 2018). For each of them, we obtained about 11 observations and monitoring was subsequently discontinued. A similar approach was followed for spectroscopic binaries, for which we acquired a number of measurements just enough to derive reliable orbital solutions (Baroch et al. 2018, 2021). For some of the binaries with the longest periods, however, monitoring at very low cadence has been extended over time to constrain better the orbital and physical parameters of the components.

Despite the discontinued targets, sometime into the survey it was realised that reaching 70 observations per star would not be possible, mostly because of the large number of measurements needed to characterise newly discovered exoplanets as a consequence of the measured astrophysical jitter and also because of the telescope and instrument overhead times being somewhat longer than initially considered. Furthermore, with the launch of the Transiting Exoplanet Survey Satellite (TESS) mission in 2018 (Ricker et al. 2015), the CARMENES Consortium agreed to invest approximately 50 useful GTO nights in following up TESS transiting planet candidates with M-dwarf hosts (CARMENES-TESS follow-up programme). As a consequence of the new circumstances, it was decided that the survey should aim at acquiring a minimum of 50 observations per target, which would yield plenty of planet detections and provide meaningful constraints on planet occurrence rates. At the same time, we redefined the relative priorities of the sample to favour stars of spectral type M4 V and later to exploit optimally the CARMENES capabilities in a relatively unexplored range of stellar host masses. Such a decision implied that the faint end of the M2 V and M3 V targets in the sample would have lower chances of being scheduled because of the employed criteria (Garcia-Piquer et al. 2017).

At the end of the GTO survey in 2020, the minimum number of 50 measurements had not been reached for all surveyed targets. The CARMENES DR1 therefore contains unequal number of observations, with a median of 30 observations per star. However, some of the targets, such as RV standards and stars with suggestive planetary signals, were observed up to a few hundred times. About two thirds of the targets have time series of at least 3 yr, and for almost half of the targets the observations cover at least 4 yr. Only 10% of the targets are observed for less than a year. The cadence is random and non-uniform, not only because of observability but also for scientific reasons and (e.g. priority increased when a planet candidate signal required more detailed sampling). In 2020, a proposal was submitted to the competitive Calar Alto Legacy projects call, and an additional 300 nights were awarded to the CARMENES Consortium to complete the survey during 2021–2023 and, hence, fulfil the goal of attaining at least 50 observations per target.

3 CARMENES GTO sample

The CARMENES GTO sample of M dwarfs is generally composed of the brightest stars of every spectral subtype that are visible from Calar Alto (δ > −23°), as described in Alonso-Floriano et al. (2015). Effectively, this means that about 70% of the full sky is observable by the CARMENES survey. We only excluded stars that are known members of visual binaries at separations closer than 5″. We explicitly did not bias our sample with regard to age, metallicity, or magnetic activity, nor did we exclude stars with planets that were already known. More information on the selection criteria was provided by Reiners et al. (2018b, hereafter Rei18b) and references therein. The sample described by Rei18b was composed of 324 stars. Throughout the survey we added nine additional targets as a result of supervening circumstances such as new exoplanet announcements, interesting targets (e.g. in the TESS continuous viewing zone), and revised spectroscopic classification. Furthermore, we added 18 targets from the CARMENES-TESS follow-up programme. As opposed to Rei18b, we also included in our current analysis double- and triple-line spectroscopic binaries and triples (SB2 and ST3, respectively) and some visual binaries. We have found 17 of such binaries in the sample, 11 of which are new additions to Rei18b, but six were present there because they had not yet been identified as SB2, ST3, or visual binaries.

Table 1 presents a selection of relevant properties of the 362 targets in the CARMENES GTO sample. The different columns list basic stellar parameters (M, R, Teff), rotation periods (Prot), and the number of measurements in the release, both in the form of pipeline-produced RVs (NRVC) and zero-point-corrected RVs (NAvc). Descriptions of these two data products are provided in Sect. 4. The basic stellar parameters were taken from the latest version of Carmencita, which is the CARMENES input catalogue (Caballero et al. 2016), and from the series of papers on the characterisation of the CARMENES GTO sample (Alonso-Floriano et al. 2015; Cortés-Contreras et al. 2017; Jeffers et al. 2018; Díez Alonso et al. 2019; Cifuentes et al. 2020; Perdelwitz et al. 2021). In the case of targets where more than one set of lines are visible in the spectra (SB2, ST3, and visual binaries), the basic parameters are not listed (as they are ill-defined) and the column NRVC provides the total number of CARMENES observations released. The penultimate column indicates if the target is part of the blind GTO survey or if it is a TESS exoplanet candidate. An asterisk marks targets already tabulated by Rei18b. We are not discussing here the statistical distribution of the target sample regarding brightness and spectral type. The general properties are equivalent to those in Figs. 2 and 3 of Rei18b, which already comprised most of the sample presented here (>90%).

The volume completeness of the CARMENES GTO sample can be investigated by comparing distance distributions with the Gaia Catalogue of Nearby Stars (GCNS; Gaia Collaboration 2021), which is assumed to be complete at the brightness cuts and spectral types of interest. In Fig. 1, we show a collection of histograms as a function of distance out to 20 pc for several spectral type and Gaia G-band absolute magnitude (MG) intervals. To allow for a comparison, spectral types of GCNS stars were estimated from MG following the corresponding relationship by Cifuentes et al. (2020). The ratio between the number of stars in the CARMENES sample and the number of known stars in the GCNS is also shown. The ratio, that is, the sample completeness, decreases with M subtype (from 27% for early Ms to 6% for late Ms), as expected due to brightness limitations. The global completeness of the CARMENES sample at 20 pc, including all spectral types, is 15%. If we consider distances to 10 pc, then the ratio of sample stars to known stars exceeds 50% in all intervals except for the latest Ms, where the ratio is 28%. Altogether, the CARMENES GTO sample contains nearly half (48%) of all known M dwarfs within 10 pc of the Sun (Reylé et al. 2021), and about 70% of those accessible from the Calar Alto Observatory. Most nearby M dwarfs that are not in the sample have close companions at less than 5″.

Table 1

Basic properties and number of measurements for the CARMENES DR1 target sample.

thumbnail Fig. 1

Distribution of the CARMENES GTO target sample (excluding the SB2 and ST3 systems) as a function of distance (d < 20 pc) for different spectral types or absolute Gaia G-band intervals. Some stars in the sample are at greater distances, and this number is provided inside the right-pointing arrow. One of the targets at the K–M spectral type boundary has an MG value below 7.73 mag and is not included; hence, the total number of stars plotted is 344. The distance distribution of the GCNS for the same intervals is also shown, and the ratios between the two are depicted as black crosses with the scale in the right y axis.

4 Observations

The observations of the CARMENES GTO survey were collected in a signal-to-noise (S/N) limited fashion. That is, using the number of counts from the exposure meter of the NIR channel (cEM) and a calibrated relationship , – the integration was continued until reaching a S/N of 150 at order 50 (~ 1200 nm) of the CARMENES NIR channel, or was interrupted after an integration of 1800 s to avoid excessive contamination from cosmic rays and line broadening due to Earth’s rotation. According to the calculations by Reiners & Zechmeister (2020), a spectrum with S/N = 150 at 1200 nm for an early- to mid-type M dwarf produces a typical uncertainty of 1 m s−1 in RV from photon shot noise, which was the required value for the survey.

A total of 19633 spectra were acquired as part of the GTO programme. However, a small fraction of them do not have sufficient quality for precise RV work and were not considered in our subsequent analysis. They were flagged by the processing pipeline because of low S/N, high S/N implying saturation risk, contamination by twilight, Moon, or stray light. The total number of spectra yielding useful RV measurements is 19 161. The discarded 472 spectra are still accessible from the Calar Alto archive2 in raw format but are not part of the CARMENES DR1.

The processing of the data was done automatically with a pipeline, including the reduction of raw frames, the extraction and calibration of spectra, the determination of RVs using a template-matching algorithm, and the calculation of cross-correlation function (CCF) products. Full details on the applied procedure are provided below. The data for SB2 and ST3 targets were only processed up to the extraction and calibration of spectra, and were not analysed to determine precise RVs because our methodology is not suitable when more than one set of stellar lines is present in the spectra. Finally, we provide the full set of data products for 18 642 out of the 19 161 good spectra.

4.1 Processing pipeline

The observations were reduced with the caracal3 pipeline, with the data flow being described by Caballero et al. (2016). The extraction pipeline is based on the reduce package of Piskunov & Valenti (2002) but many routines have been revised. In particular, we developed the flat-relative optimal extraction (FOX, Zechmeister et al. 2014) and wavelength calibration scripts, which combine spectra from hollow-cathode lamps (HCLs) and Fabry-Pérot (F-P) étalons (Bauer et al. 2015). The data release in this work is based on caracal v2.2S.

4.2 Radial velocities

The RVs for the CARMENES DR1 were computed with serval4 (Zechmeister et al. 2018) and raccoon5 (Lafarga et al. 2020). Both software packages were specifically developed for data coming from the CARMENES instrument, although they can process spectroscopic data from other precise RV instruments as well (e.g. Stefánsson et al. 2020; Hoyer et al. 2021; Wang et al. 2022; Turtelboom et al. 2022).

The serval code implements a data-driven approach, where both RVs and templates are derived from the observations themselves via a least-squares fitting procedure similar to the Template-Enhanced Radial velocity Re-analysis Application (TERRA; Anglada-Escudé & Butler 2012). The co-adding is performed by cubic B-spline regression. For the barycentric correction, the default option is the Python implementation barycorrpy6 (Wright & Eastman 2014; Kanodia & Wright 2018). The RVs for each spectral order are produced, and the global RV of the spectrum is subsequently computed as a simple weighted mean over the spectral orders. By default, the ten bluest and the ten reddest spectral orders are not used. In those regions, the instrument efficiency decreases. Furthermore, the red end is strongly affected by telluric contamination and dichroic cutoff. For faint late M dwarfs additional blue orders may be omitted because of low S/N. Since the present data release contains the order-wise RVs, a more sophisticated recalculation of RV values (robust means, re-weighting using a posteriori information) employing a detailed chromatic analysis is also possible (e.g. Zechmeister et al. 2019). Finally, corrections for instrumental drift and secular acceleration (Kürster et al. 2003) are applied to the global RV, yielding the so-called RVC (Radial Velocity Corrected) velocities.

The RV error bar is calculated as the weighted mean of the order-wise RVs (see Eq. (15) in Zechmeister et al. 2018) and takes into account photon noise, readout noise, and model mismatch. The last contribution quantifies the difference between the spectrum and the template (caused, for example, by cosmic rays, telluric contamination or detector artefacts), and the excess scatter of the averaged individual orders (caused, for example, by telluric contamination affecting specific orders or a chromatic trend). Thus, formal RV uncertainties are based on the quality of the template fit and not on estimates of any physical effects during observation or calibration (e.g. modal noise). The CARMENES instrument was designed to minimise all such effects (Seifert et al. 2012; Stürmer et al. 2014) but, if anyway present to some extent, they will result in excess noise (instrumental jitter).

A further RV data product is provided, namely AVC (Average Velocity Corrected) velocities. These are obtained from RVCs by correcting for nightly zero points (NZPs; see Sect. 4.4). AVC RVs are not calculated if no instrumental drift value is available. The total error bar of each AVC RV considers the uncertainties of the RVC and the corrections added in quadrature. In addition to RVs, serval provides a further set of useful parameters. These include the chromatic index (CRX; a measure of the wavelength dependence of the RVs), the differential line width (dLW), and spectral line indices (e.g. Hα, Ca I, Ca II IRT (infrared triplet), and Na I D), which are valuable activity indicators (Fuhrmeister et al. 2019b; Schöfer et al. 2019). A full description of these serval products and their calculation methodologies is provided in Zechmeister et al. (2018).

The raccoon code is based on the CCF concept (Baranne et al. 1996), whereby a weighted binary mask is used to calculate the convolution with each observed spectrum. In our implementation, we derived the mask from the serval template of the target itself. One of the outputs is the RVs, which are known to be less precise than values coming from template matching for M dwarfs (Perger et al. 2017), but still allow for a crosscheck with serval. Other relevant CCF parameters produced by raccoon are the contrast (CON), the full width at half maximum (FWHM), and the bisector inverse slope (BIS). These parameters can be regarded as moments of the CCF that carry information on the characteristics of the stellar lines and, therefore, can be used to assess variability coming from astrophysical sources. Further details can be found in Lafarga et al. (2020).

The RVs in the CARMENES DR1 may differ from RVs that have appeared in previous CARMENES publications. This is because serval and raccoon are steadily maintained and new upgrades are continuously made. In addition, all parameters are recalculated when new spectra of a target are considered (i.e. producing a new template) and, thus, slightly different values may result. Finally, the NZP corrections can vary when new data are considered, also impacting on the final velocities. In any case, any differences with published data are generally minor.

thumbnail Fig. 2

Nightly zero points (NZPs) of CARMENES VIS. The RVs of all RV constant stars (small green points) are zero-centred. A mean (the NZP, solid black circles) is computed for each night, in which identified outliers (small red points) are omitted. NZPs with fewer than three RV standard observations (open magenta circles) are replaced by a local mean.

4.3 Telluric contamination correction

The Earth atmosphere imprints spectral features from its molecular and atomic components (mostly H2O and O2 in the VIS domain), called telluric lines (or tellurics, for short), onto the stellar spectrum. serval handles this contamination by simply masking telluric lines when computing the RVs (during co-adding, telluric lines are strongly down-weighted, and severely contaminated template regions are masked as well during RV computation). Masking lines is a straightforward and robust first-order approach. The default mask of serval flags regions where the telluric line depth is typically about 5% or greater. Various tests using different thresholds and resulting mask widths showed this value to provide optimal results by trading off wavelength coverage (i.e. RV precision) and systematic effects from telluric contamination. While the telluric mask is static in the detector frame, it moves in the stellar rest frame because of Earth’s yearly barycentric motion. To ensure that identical spectral regions are used for RV determination throughout the observing season, an alternative approach would be to mask out the full barycentric velocity range around each telluric feature. However, we preferred not to use such a procedure because it significantly diminishes the available wavelength range and, thus, the amount of RV information.

There may be cases where the residual telluric RV content (due to high airmass or micro-telluric contamination) may still be significant. Such residuals can most likely affect cases where the RV internal precision is very high (e.g. high S/N observations) and where the stellar RV signal is weak (e.g. fast rotators). Residual telluric contamination can result in spurious RV periodicities, mostly yearly signals or their aliases (Damasso et al. 2022). Hence, caution is advised in the interpretation of those typically long-period, low-amplitude signals. Improvements may be made by re-weighting spectral orders, reprocessing with more conservative masks, or employing a more sophisticated telluric modelling scheme (e.g. Nagel 2019).

4.4 Nightly zero points

Although the CARMENES spectrograph is usually wavelength calibrated each afternoon and nightly instrumental drifts are measured with the F-P étalon, stellar RVs from the same night often share common systematic effects, which produce NZP offsets generally of a few m s−1 with a median error bar of 0.9 m s−1 (see Fig. 2). We employ RV-constant stars (rms < 10 m s−1) to calculate NZPs, with the exact procedure being described in more detail by Trifonov et al. (2018). The resulting values are subsequently subtracted from each of the serval RV measurements. To avoid self-biasing the measurements, the zero point of RV-constant stars is calculated by removing the target itself from the calibration pool (Tal-Or et al. 2019). Tests revealed that NZP-corrected RVs improve the statistical significance of the best-fit models of CARMENES exoplanet discoveries, thus illustrating the benefits of the correction procedure. The same algorithm was applied by Tal-Or et al. (2019) to archival HIRES (High Resolution Echelle Spectrometer) Keck RVs and by Trifonov et al. (2020b) to reprocessed RVs from HARPS (High Accuracy Radial velocity Planet Searcher) spectra. In both cases the studies revealed and corrected systematic effects in those instruments.

Table 2 provides NZP values for all the CARMENES GTO nights. Reasons explaining the nightly offsets can be various, including a drift of the F-P, degraded quality of aged HCLs, strong instrument drifts during the ~ 15 min calibration sequence (F-P and HCL calibration frames cannot be taken simultaneously), and different injection of calibration light coupled with insufficient scrambling. We were able to reduce some fraction of the night-to-night variability found during the initial CARMENES operations through hardware configuration changes and by employing a different strategy when acquiring the daily calibration sequences. As a result, the NZP scatter diminishes slightly after 2 yr of operation (Fig. 2). In addition to the night-to-night offsets, we also performed a correction for intra-night drift. The correction was found to be significant early in the survey and related to a temperature effect of the F-P subsystem. A hardware upgrade on 6 September 2017 (BJD 2458003) greatly decreased the temperature coupling and eliminated the need for such a correction. In any case, the effects of self-bias and intra-night drift correction are small. Further details on the instrument performance are provided in Bauer et al. (2020).

Table 2

Nightly zero points for CARMENES.

5 Results

The CARMENES DR1 provides raw spectroscopic data for the total sample of 362 targets but only full data products (including RVs, spectroscopic indices, and CCF parameters) for 345 targets, that is, excluding 17 SB2 and ST3 systems. Precise RVs of the components of 12 of these spectroscopic multiple systems and full orbital and physical analyses were presented by Baroch et al. (2018, 2021). The remaining five binary systems, namely J05084−210, J06396−210, J09133+668, J16343+571 (CM Dra), and J23113+085, do not have a publication using CARMENES data yet. The procedure described in Sect. 4 was applied to the 18 642 suitable spectra and these produced the same number of RV determinations and associated data products. However, a fraction of those measurements lack a velocity drift calculation because of the poor quality of the simultaneous F-P spectrum. As a consequence, the number of drift-corrected RV measurements is 17 749, and these correspond to 344 targets. Only the faint target J16102−193 (K2–33) is not in the final sample because all spectra were taken without simultaneous F-P. All data products associated with the CARMENES DRl and ancillary files are available online7.

In Fig. 3, we illustrate the distribution of the formal uncertainties of the RV measurements (internal precision). Targets are grouped into four spectral-type bins using the same criteria as in Fig. 1. Brighter targets have typical uncertainties of ~1 m s−1, as their S/N at 1200 nm reached 150, but fainter targets have larger uncertainties due to the larger photon noise. The median value of the internal precision is 1.27 m s−1, with the maximum of the distribution (mode) at 0.91 m s−1.

The distribution of observations and their dispersion are illustrated in Fig. 4, also grouped in spectral-type bins. The scatter plot depicts the rms of the RV time series for each of the 344 NZP-corrected targets as a function of the number of observations, Nobs Histograms of rms and number of observations are shown along both axes of the plot. The weighted rms of n = 1…Nk observations of each target k is calculated as

(1)

where the epoch weights wn include a jitter term σj, which is added in quadrature to the formal RV uncertainties σRV,n

(2)

The σj and the re-weighted mean RVs were obtained self-consistently for each star via a maximum likelihood optimisation8. For the calculation we used the RVs as measured, with NZP correction, and no known signals of any nature (activity, planets) were subtracted. The median and mode of the distributions are 3.9 m s−1 and 3.3 m s−1, respectively. These values can be compared to those characterising the internal precision in Fig. 3 to conclude that the RVs are most likely dominated by jitter (and signal) from astrophysical sources, which is statistically estimated to have a median contribution of ~3.5 m s−1. No obvious rms trends as a function of spectral type are observed except for a much higher rms (27.5 m s−1) for the latest bin due to a large fraction of low S/N measurements.

thumbnail Fig. 3

Stacked histograms and statistical parameters (median and mode) of the internal precision (formal uncertainties) of the 18 642 precise RV measurements in the CARMENES DR1.

thumbnail Fig. 4

Distribution of observations. The rms is computed for each of the 344 targets (circles colour-coded by spectral type) from their serval RV time series (NZP-corrected). No periodic signals (activity or planetary) were removed. The faint target J16102−193 (K2–33) is excluded, since all its spectra were taken without F-P.

5.1 High-resolution spectroscopic time series data

We compiled data tables of the time series of the RVs as described in Sect. 4, as well as additional ancillary parameters, such as stellar activity indices, for each of the 345 M dwarfs (excluding SB2 and ST3) in the CARMENES sample. The dataset includes RV, CRX, dLW, and chromospheric line indices (Hα, Ca I, Ca II IRT, and Na I D) from serval, and the CCF RV, BIS, FWHM, and CON obtained with raccoon. Data from each spectral order are provided separately. For the RVs, both the values produced by serval (RVC) and those obtained after applying NZP corrections (AVC) are included. Furthermore, the exposure time and airmass of the observations, and the instrumental drift, the barycentric Earth RV, and the secular acceleration corrections applied to calculate RVCs are also provided.

Graphical representations of the time series of serval RVs corrected for NZPs as described in Sect. 4.4 have been produced for all targets with at least five valid NZP-corrected RV values and are available online7. An example is provided in Fig. 5 for the target J00051+457 (GJ 2). Periodogram analyses of the RVs and several relevant activity indices are also presented. Before computing the periodogram, we applied a clipping criterion to the measured values of the RVs and indices to avoid obvious outliers and poor-quality measurements. All data points deviating by more than 3σ from the mean were eliminated and so were measurements with error bars greater than the average value plus 3σ. Nightly averages were computed for the targets J00183+440 (GX And), J00184+440 (GQ And), and J07274+052 (Luyten’s Star), as they had observations at higher cadence.

We subsequently computed the generalised Lomb-Scargle (GLS) periodogram (Zechmeister & Kürster 2009) of the RVs, the CRX, dLW, Ha, and Ca II IRT indices, and the CCF parameters BIS and FWHM. We grouped the indices and CCF parameters in pairs according to their expected sensitivity to the same activity phenomena, according to the analysis of Lafarga et al. (2021). Therefore, three panels with equivalent activity indicators are provided in Fig. 5, namely CRX & BIS, dLW & FWHM, and Hα & Ca II IRT. We considered periods ranging from twice the time span of the observations (to identify long-term variations) to the Nyquist frequency as computed from the closest RV measurement pairs of each dataset. However, an upper frequency limit of 0.95 day−1 was set to avoid daily aliases, except for targets with known short-period periodicities (close-in transiting planets and very fast-rotating stars). We calculated the false alarm probability (FAP) by running 105 bootstrap realisations of the datasets. From the bootstrapped data, we also computed the probability of each periodogram peak by assessing the number of times that the real periodogram at a given frequency is above all the realisations. This probability is related to the GLS power (Zechmeister et al. 2009) and is used in the graphical representation.

thumbnail Fig. 5

Time series data and periodograms for target J00051+457 (GJ 2). Top panel: NZP-corrected RV time series. Outlier measurements (grey symbols; see text) are excluded in the periodogram calculation. Bottom panels (from top to bottom): GLS periodograms of the NZP-corrected RVs (AVC, blue), the chromatic index (CRX, black) and CCF bisector span (BIS, red), the differential line width (dLW, black) and the CCF FWHM (red), and the chromospheric activity indices corresponding to the Ha (black) and averaged Ca II IRT lines (red). The dash-dotted blue vertical line in all panels marks the position of the most significant peak in the RV periodogram, while the circles in the sub-panels highlight the position of the strongest signal in each periodogram, with the period given in the legend. The 0.1%, 1%, and 10% FAP levels are shown as horizontal dashed, dash-dotted, and dotted green lines, respectively, and are visible if they fall within the displayed vertical range. Vertical axes in panels that show two different datasets are scaled such that the FAP levels are identical. The colour code allows the graph to be related with the axis labels and legends.

5.2 Exoplanets in the CARMENES GTO sample

The CARMENES sample was designed to preserve completeness as much as possible. Therefore, the initial target selection did not explicitly exclude known planet hosts. The inaugural CARMENES survey paper by Trifonov et al. (2018) analysed the CARMENES data for a sample of seven targets known to host 12 planets. In this study, a hitherto unknown second, long-period planet orbiting J11417+427 (GJ 1148) was reported, qualifying as the first exoplanet discovered by CARMENES. Shortly after, Reiners et al. (2018a) published the first exoplanet detected from data collected solely from the CARMENES survey. Since then, a succession of announcements has been made using data from the CARMENES blind survey, totalling 33 newly discovered planets in 28 planetary systems at the time of writing this paper. The new CARMENES planets are marked with a‘d’ in column ‘Type’ of Table A.1. In some cases, CARMENES data were combined with precise RVs from other instruments (such as HARPS, HARPS-N, ESPRESSO, HIRES, IRD, MAROON-X, etc.) to enhance the statistical significance of the measurements.

Moreover, over the course of the survey, five already announced exoplanets were re-analysed using CARMENES data. Together with the 12 known planets in Trifonov et al. (2018), this makes up a total of 17 planets that are marked with an ‘r’ in Table A.1.

As explained in Sect. 1, the CARMENES Consortium decided to invest a fraction of the GTO time in following up on transiting planet candidates. Some of the targets came from the K2 mission (Howell et al. 2014), but most of them were provided by the ongoing TESS mission. The campaign has been fruitful, and CARMENES has led or contributed to the confirmation of 26 such planet candidates and helped measure their masses. The CARMENES planets resulting from follow-up activities are marked with an ‘f’ in Table A.1. The TESS planet candidate around J11044+304 (TOI-1806) has been followed up and validated with CARMENES; however, its parameters are not listed in the table because they are not yet sufficiently significant.

The columns in Table A.1 provide basic information on the targets and their planets. The parameters are taken from each of the quoted references. NCAR and Nother are the number of RVs from CARMENES and other instruments, respectively, that were used in the corresponding publication. NCAR may differ from the number of measurements in the DR1 release. Cases where NCAR is greater than the number in DR1 correspond to recent publications that include observations taken after 31 December 2020, as part of the new CARMENES Legacy+ survey, while cases where NCAR is below the number of measurements in DR1 are those where additional measurements within the CARMENES GTO were taken after the quoted publications. For four such planets we present new parameters in Table A.1 considering all the measurements in DR1. The four revised planets are J06548+332b (GJ 251 b; Stock et al. 2020b) J08413+594b,c (GJ 3512b,c; Morales et al. 2019), and J16167+672S b (HD 147379 b; Reiners et al. 2018a).

In addition to the publications using CARMENES RVs, the DR1 includes measurements of targets for which there have been exoplanet detections or claims in the literature but do not have a specific publication using CARMENES data at the time of writing. These are listed in Table 3, along with the number of released CARMENES epochs. Besides planet detections and confirmations, there are some targets in our sample for which planets have been announced and are listed in exo-planet catalogues but could not be confirmed or are controversial given the data obtained with CARMENES or other instruments. The list of such planets is provided in Table 4. We are not including a planet around J00183+440 (GX And b) because the CARMENES observations now seem to support a planetary scenario for the 11.44-day signal (Trifonov et al., in prep.), in contrast to the initial CARMENES data (Trifonov et al. 2018), which were casting doubt on its nature.

Some of the challenged planets were already discussed in dedicated publications, as listed in Table 4. In two other cases, we carried out the analysis as part of the present work. Particularly, the candidates announced around J02222+478 (GJ 96) and J09561+627 (GJ 373) can be quite confidently ruled out as planets. For GJ 96, Hobson et al. (2018) announced a planet candidate based on 72 SOPHIE RVs9. The periodograms in Fig. 6 show that the 75-day signal present in SOPHIE data is absent in the 53 RVs from CARMENES. Instead, the dominant signal is at 28.5 days, which Hobson et al. (2018) already attributed to stellar activity. Indeed, this period is also present in the dLW, Ca II IRT, and Ha time series of the CARMENES data, thus ruling out its planetary nature. Phase-folded plots are shown in Fig. 7. The signal in GJ 373 at about 17.8 days announced as a planet by Tuomi et al. (2019) and Feng et al. (2020) can most definitely be attributed to stellar rotation modulation since it appears strongly in CARMENES activity indicators such as dLW and Hα and some CCF parameters, as can be seen in Fig. 8.

The sample of planets in Table A.1 is represented graphically in Fig. 9. Scatter plots combine the stellar mass, minimum planetary mass, orbital period, and RV semi-amplitude. The planet samples in the diagrams comprise those with CARMENES analyses (showing ‘d’, ‘r’, and ‘f’ separately in Table A.1) and those coming from the NASA Exoplanet Archive10. The latter correspond only to RV-detected planets (i.e. planets discovered through photometric transits are excluded). Also, histogram distributions of each of these quantities for the planets analysed with CARMENES data are depicted as side plots.

A few features in Fig. 9 are worth discussing. Regarding stellar mass, a majority of CARMENES planets have host stars with masses between 0.25 M and 0.65 M, which constitute the bulk of the sample. Remarkably, half of the 24 RV planets with stellar hosts below 0.25 M known to date have been discovered by CARMENES, a testament of the advantage offered by a red-optimised RV spectrometer in the late-type host regime. In terms of planetary mass, the majority of the CARMENES planets are in the super-Earth to the Neptune-mass domain, although several Earth-mass planets have been detected orbiting some of the lower-mass targets in our sample. Remarkable cases are two systems, J02530+168 (Teegarden’s Star; Zechmeister et al. 2019) and J00067−075 (GJ 1002; Suárez Mascareño et al. 2023), each with two Earth-mass planets within the liquid-water habitable zones of their stars. Also, CARMENES has discovered six Saturn- and Jupiter-mass planets, some of them around very low-mass primaries, thus defying canonical planet formation models (Morales et al. 2019), which predict very low occurrence rates of giant planets around M-type dwarfs (e.g. Schlecker et al. 2022). As expected for detectability reasons, most of the CARMENES planets have orbital periods from a few days to a few tens of days. Although not discussed here, the CARMENES GTO survey also announced (Baroch et al. 2021) two brown dwarf candidates on very long-period orbits (P ≳ 3000 days), around 10504+331 (GJ 3626) and J23556−061 (GJ 912).

Table 3

CARMENES survey targets with confirmed or claimed exoplanets that have no dedicated publications with CARMENES data.

Table 4

Exoplanets challenged by CARMENES.

thumbnail Fig. 6

Same as Fig. 5, but for J02222+478 (GJ 96). No planetary signal at a period of 73.9 days, as claimed by Hobson et al. (2018), is visible in the CARMENES data. The only significant periodicity is at ~28.5 days and seems to be related to activity given the counterparts in some of the indicators. Phase-folded plots are shown in Fig. 7.

thumbnail Fig. 7

Phase-folded RV data of J02222+478 (GJ 96). Left. SOPHIE (green squares) and CARMENES VIS (blue circles) phase-folded to a period of 73.94 days. The planetary Keplerian signal (dashed green line) proposed by Hobson et al. (2018) is based on the SOPHIE data alone and disfavoured by the CARMENES measurements. Right same as the left panel, but phase-folded to the periodicity of 28.3 days, which is most likely associated with stellar activity.

5.3 Planet occurrence rates

Using the CARMENES DRl sample we calculated planet occurrence rates in a similar way as was done by Sabotta et al. (2021, hereafter Sab21). In that work, preliminary occurrence statistics were calculated using a subsample of 71 targets having at least 50 CARMENES RV measurements. The re-analysis in the present work applies similar target selection criteria. From the initial 362 targets in the CARMENES GTO sample we excluded 124 targets because of several possible reasons: (i) they were added later for transit follow-up (mostly TES S candidates; 20 targets); (ii) they are spectroscopic binaries and triples (23 targets); (iii) they are part of the RV-loud sample as defined by Tal-Or et al. (2018) (52 targets); or (iv) we obtained fewer than ten RV measurements (29 targets). The sample therefore comprises a total of 238 targets, including 69 of the 71 targets in Sab2l. For the two targets not included in the CARMENES DRl, one was excluded after being classified as a late-K dwarf (J18198−019, HD 168442), and the other one was subsequently classified as a resolved binary (J23113+085, NLTT 56083).

For the planetary sample, we re-ran the signal retrieval and vetting algorithm from Sab2l (see the results in Table A.2). The only change that we made was the period limit used for the long period planets. Sab2l included every signal if the time baseline was longer than two orbital periods, while here we include every signal with time coverage of at least 1.5 times the orbital period. If we considered the more conservative period limit of Sab2l, we would exclude several giant planets from the planet sample and that would therefore reduce the statistical soundness of our analysis. As a result, we regard the new criterion as a better balance between being too conservative but still making sure that the signal is indeed periodic. Using this criterion, we identified 37 planets that can be confirmed with CARMENES data alone and three additional planet candidates (around J05033−173, J17033+514, and J18409−133). We also include 13 planets with fewer than 50 RVs that were detected using data from other surveys (mainly HADES, HArps-n red Dwarf Exoplanet Survey, Pinamonti et al. 2022, and HARPS, Bonfils et al. 2013) if they induce an RV semi-amplitude of K > 2 m s−1. We assume that we would have detected such planets if we had not terminated the observations because of our independent knowledge. We obtained those targets from a comparison with the two exoplanet databases on The Extrasolar Planets Encyclopaedia11 (Schneider et al. 2011) and the NASA Exoplanet Archive10. In Table A.2, we mark planets that are listed in one of the databases and are well below our detection limits, planets with fewer than 50 RVs that are included in our planet sample, and archive planets that are not supported by CARMENES data. In this way we increase the planetary sample in Sab2l by 26 planets (from 27 to 53) for the recalculation of the occurrence rates. The total number of 53 planets reside in 43 planetary systems.

We calculated individual planet detection maps for all targets following the procedure described by Sab2l. The numerical and graphical results are available online7. The global detection probabilities across the period-mass plane and the planets mentioned above are shown in Fig. 10, with the colour scale indicating the average of all detection probabilities for the individual grid points. There are five planets in a low-probability region, which means that we can only detect such planets for a small fraction of our sample. Unsurprisingly, these are Teegarden’s Star b and c, YZ Cet c and d, and Wolf 1069 b, all of which are Earth-mass planets with very low-mass stellar hosts.

Using the same method as in Sab2l, we obtained the power-law distribution in Mpl sin i for the occurrence rate estimate (Fig. 6 in Sab2l). The updated power-law with Npl, corrected = a (Mpl sin i)α is only slightly shallower, with the slope changing from α = −1.14 ± 0.16 to α = −1.05 ± 0.01 for planets with masses below 30 M, and from α = −0.26 ± 0.17 to α = −0.14 ± 0.25 for higher-mass planets (see Fig. 11). For our occurrence rate determination, we used this power-law as an initial assumption on the Mpl sin i distribution, instead of assuming a log-uniform distribution12. The results are summarised in Table 5. We report both the number of planets per star () and the frequency of stars with planets (Fh). To obtain the latter, we repeated the analysis but instead of including all planets, we reduced the planet sample and took only the single planet with the highest K amplitude in the system. We then inspected the complete period-mass plane with periods of 1 day to 1000 days and Mpl sin i of 1 M to 1000 M. In this parameter range, we determined an overall occurrence rate of planets per star and stars with planets. This means that the planet multiplicity in our sample is around 1.5 planets per system.

The analysis of Sab2l yielded occurrence rates that are larger by a factor of two for planets with 10 M < Mpl sin i < 100 M and by 30% for the low-mass planets with 1 M < Mpl sin i < 10 M with respect to the results obtained here for the full sample. The lower occurrence rates cannot be due to the looser requirement on orbital period coverage (1.5 instead of two orbital periods), since, if anything, this would result in larger occurrence rates. The smaller occurrence rates observed for the full CARMENES sample thus illustrate the effectiveness in terms of planet discovery of the pre-selection of targets that are observed more intensively than others. Human intervention bias in this case leads to an over-estimation of occurrence rates. The survey sensitivity is higher for stars with planets because targets showing interesting signals that could be of planetary nature were observed more intensively. Sab21 pointed out this effect, and explicitly introduced the bias by rejecting all targets with fewer than 50 RVs, but it affects all targeted surveys that change the observing strategy based on acquired knowledge. In fact, by aiming at a specific number of observations for all of our targets, we minimised this effect. In the CARMENES DR1, we reach this number of 50 RVs for 42% of our targets, which corresponds to 112 stars. We are continuing the survey as part of the CARMENES Legacy+ programme. Even if the planet detection efficiency may not be as high as in the early stages of the GTO, the statistical value of the sample will greatly increase.

We compare our low-mass planet occurrence rates around M dwarfs to those of other surveys in Fig. 12. Our updated occurrence rates are consistent with the values obtained from the HARPS (Bonfils et al. 2013) and HADES (Pinamonti et al. 2022) surveys, but our results are based on a significantly larger statistical sample. The agreement is good despite the fact that both estimates by Bonfils et al. (2013) and Pinamonti et al. (2022) assumed a log-uniform distribution in planet mass, as opposed to our power-law relationship. If we also utilised a uniform distribution for our occurrence rate calculation, we would have obtained a lower occurrence rate of low-mass planets per star in orbits of up to 100 days (indicated as the grey square in Fig. 12).

In this parameter range, we obtained instead 1.06 ± 0.17 planets per star assuming a power-law distribution in Mpl sin i. The difference occurs only in those regions of the period-mass plane with a strong sensitivity gradient, that is, below 10 M. For higher planet masses, the choice of distribution does not affect our results significantly.

The comparison to transit surveys is not as straightforward due to the lack of an exact correspondence between the observed parameters. The expected value of sin i assuming randomly oriented orbits is ~0.8 (e.g. Hatzes 2019) and, therefore, our Mpl sin i bin of 1–10 M on average corresponds to a bin of 1.25– 12.5 M in true Mpl. In this mass regime, planets could be rocky, water worlds, or puffy sub-Neptunes with very different densities (Luque & Pallé 2022). According to the mass-radius relation of Kanodia et al. (2019), this mass interval corresponds on average to the Rpl interval of 1.3-3.7 R. In log-space this is only 75% of the radius interval of 1–4 R that Sab21 used for comparison with transiting planet statistics. Thus, in Fig. 12, we plot lower occurrence rates for the transit surveys (75% of those in Sab21). In any case, all occurrence rate estimates agree within a factor of two despite all the involved assumptions and the fact that we infer the occurrence rates from an overall detection sensitivity of 15% (considering the full period-mass plane).

The discussion above is relevant if one wants to find an absolute number of planets per star or to compare with transiting planet surveys or surveys targeting other stellar masses. Moreover, these calculations also serve as a valuable benchmark for planet formation theories that aim to reproduce populationlevel trends of exoplanets (e.g. Ida & Lin 2004; Bitsch et al. 2015; Miguel et al. 2020; Izidoro et al. 2021; Schlecker et al. 2021a,b; Mishra et al. 2021). Using the results of Sab21 as input, Schlecker et al. (2022) compared a planet sample based on the HARPS and CARMENES M-dwarf surveys to a synthetic population computed with the Bern model of planet formation (Mordasini et al. 2012; Emsenhuber et al. 2021; Burn et al. 2021) and found three main discrepancies.

The first one is the observational finding of an excess of giant planets around lower-mass stars compared to the theoretical prediction. The simulations do not produce any giant planets around host stars with masses below 0.5 M. As was done by Sab2l, we split the full CARMENES sample at a stellar mass of 0.337 M and calculated giant planet occurrence rates. The median stellar masses of the two subsamples are 0.24 M and 0.45 M. Using a strict limit for the giant planet mass of Mpl > 100 M, we obtained a rate of planets per star and planets per star for the low-mass and the high-mass stellar subsamples, respectively. The resulting occurrence rate ratio, fhighmass/flowmass = 2.14, is marginally consistent with the giant planet frequency as a function of stellar mass published bv Ghezzi et al. (2018).

(3)

assuming similar stellar metallicities, [Fe/H], in both samples.

The second discrepancy between model and observation concerns the shape of the planet mass distribution. The distribution of Mpl sin i in the synthetic population is bimodal, whereas its counterpart in the observed sample is consistent with a power-law. In fact, our planet mass distribution does not deviate significantly from that in Sab2l (see Fig. 6 therein).

A third mismatch between the observed and model-predicted planet demographics as identified by Schlecker et al. (2022) is the orbital period distribution around stars with masses higher than 0.4 M. Short-period planets (Porb < 10 days) are found to be significantly scarcer in the observed population compared to the synthetic one. The drop in occurrence rates at periods of less than 10 days, which was previously observed for stars with different stellar masses, does not hold for targets with masses below 0.4 M, with period distributions showing a good match. One possible explanation is a migration barrier having higher efficiency in protoplanetary disks around early M dwarfs that is not adequately accounted for by current models. For targets with M < 0.337 M we calculate and low-mass planets per star for the intervals 1–10 days and 10–100 days, respectively.

thumbnail Fig. 8

Same as Fig. 5, but for J09561+627 (GJ 373). The periodic signal at 17.8 days attributed to a planet by Tuomi et al. (2019) and Feng et al. (2020) most likely arises from stellar activity in view of the counterparts in most of the activity indicators.

thumbnail Fig. 9

Scatter plots of the CARMENES DR1 exoplanet sample compared to the complete sample of catalogued planets in the NASA Exoplanet Archive detected via RVs (903; small dots). Different symbols indicate planets newly detected from the CARMENES blind survey (33; stars), planets confirmed from transit follow-up (26; circles), and known planets re-analysed with CARMENES data (17; triangles). The three panels correspond to pairs of different relevant parameters, with the complementary colour scale introducing a third dimension. The histograms along the axes show distributions of the corresponding parameters for the CARMENES planet sample. The blue shaded band in the top-right panel represents the liquid-water habitable zone with limits defined by the ‘runaway greenhouse’ and ‘maximum greenhouse’ criteria (Kopparapu et al. 2013).

thumbnail Fig. 10

CARMENES detection probability map in the period-mass plane derived from the observations of 238 targets. The detection probability is calculated using each target mass, and the average of all detection probabilities for the individual grid points is shown in the period-planet mass plane as a colour map. For example, a 10% detection probability for 238 targets indicates that the survey data can detect the respective planet in approximately 24 targets. Yellow circles indicate CARMENES planet detections and planet detections from other instruments that are confirmed by CARMENES data; grey circles planet candidates; and red circles planet detections from other surveys for which no sufficient CARMENES measurements are available to confirm them. The five planets in the low-probability detection region (<10%) are Teegarden’s Star b and c, YZ Cet c and d, and Wolf 1069 b, all of which are Earth-mass planets with host stars of less than 0.2 M, thus suggesting that such planets are very abundant.

thumbnail Fig. 11

Histogram of the number of planets detected in six Mpl sin i bins. The squares indicate the number of planets corrected for the survey sensitivity averaged over periods of up to 240 days. The dashed line and the solid line indicate the power law fit for Mpl < 32 M and Mpl > 32 M. respectively.

Table 5

Planet occurrence rates for a sample of 238 M dwarfs in the CARMENES DR1 sample, including candidates and planet detections from other surveys.

6 Conclusions

The CARMENES GTO survey ran from 1 January 2016 to 31 December 2020 and obtained 19633 spectroscopic measurements of a sample of 362 targets. The sample was designed to be as complete as possible by including M dwarfs observable from the Calar Alto Observatory with no selection criteria other than brightness limits and visual binarity restrictions. To best exploit the capabilities of the instrument, variable brightness cuts were applied as a function of spectral type to increase the presence of late-type targets. This effectively leads to a sample that does not deviate significantly from a volume-limited one for each spectral type. The global completeness of the sample is 15% of all known M dwarfs out to a distance of 20 pc and 48% at 10 pc.

The present paper accompanies the release of a large dataset, the CARMENES DR1. Raw data, pipeline-processed data, and high-level data products are provided, including 18 642 precise RVs for 345 targets (removing double- and triple-line systems). After correction of a NZP offset, the median internal precision of early and intermediate M-dwarf types is ~1.2 m s−1. This value increases to ~5.4 m s−1 for late M spectral types due to their intrinsic faintness. The median rms of the RV time series of all the targets in the sample is calculated to be ~3.9 m s−1, where no signal has been subtracted. A comparison between the internal and external precisions indicates that the RV variance has a contribution of ~3.5 m s−1 on top of the instrument error when treated as uncorrelated random noise. This external noise component is unlikely to be of instrumental origin. It is instead believed to arise from astrophysical effects, including Keplerian signals from planets but, most importantly, RV variability arising from stellar activity (e.g. active region rotation and evolution).

The CARMENES time series data have been analysed in the search for RV signals of a planetary nature. So far we have identified 33 new planets from the blind survey observations, which are complemented by 17 planets that we have re-analysed with CARMENES data and 26 planets from transit search space missions that we have confirmed and measured. The number of blind survey planets is in good agreement with the initial estimates considering the properties of the stellar sample, the survey design, and the assumed planet occurrence rates (Garcia-Piquer et al. 2017). The new planets cover a broad region of the parameter space in terms of stellar host mass, planetary mass, and orbital period. A remarkable result is that CARMENES has discovered half of the RV planets known to orbit stars of masses below 0.25 M. This fact illustrates the prime ‘hunting ground’ of CARMENES thanks to the competitive advantage of the optimised red-sensitive design and the possibility of undertaking a massive survey with a large fraction of dedicated 4 m class telescope time over 5 yr.

With the CARMENES DR1 data, we have calculated new planet occurrence rates around M dwarfs to update the results already presented by Sab21. We have employed a subsample of 238 stars that fulfil a set of specific requirements. We still find a high long-period giant planet occurrence rate of around 3%, a high number of low-mass planets (1.06 planets per star in periods of 1 day to 100 days), and an overabundance of short-period planets around the lowest-mass stars of our sample compared to stars with higher masses. For our complete periodmass parameter space, we determine an overall occurrence rate of planets per star and a fraction of stars with planets. We calculate the overall CARMENES survey sensitivity to be 15% and find planets around 43 of 238 targets (i.e. 18% of the stars), which again shows that nearly every M dwarf hosts at least one planet.

In the present description of the CARMENES GTO data, we have focused on their use for precise RV work in the field of exoplanet detection and characterisation. Nevertheless, we have shown in a number of publications that these data are also of high value to a variety of science cases within stellar astrophysics, such as studying atmospheric parameters (Teff, log ɡ, and chemical abundances; Passegger et al. 2018, 2019, 2020, 2022; Fuhrmeister et al. 2019a; Marfil et al. 2020, 2021; Abia et al. 2020; Shan et al. 2021), determining fundamental properties (M, R, and magnetic field; Schweitzer et al. 2019; Shulyak et al. 2019; Reiners et al. 2022), and analysing magnetic activity (Tal-Or et al. 2018; Fuhrmeister et al. 2018, 2019b, 2020, 2022; Schöfer et al. 2019; Hintz et al. 2019, 2020; Baroch et al. 2020; Lafarga et al. 2021; Jeffers et al. 2022). CARMENES VIS channel data have also proved useful in addressing the study of exoplanet atmospheres via transit transmission spectroscopy (Yan et al. 2019, 2021; Casasayas-Barris et al. 2020, 2021; Sánchez-López et al. 2020; Khalafinejad et al. 2021; Czesla et al. 2022) and the Rossiter-McLaughlin effect (Oshagh et al. 2020; Sedaghati et al. 2022).

The CARMENES GTO survey is now complete. In terms of exoplanet RV detection, the survey has provided about 60 planet discoveries and confirmations, some of which are of very high scientific relevance, and, as a sample, is of great statistical value, thus contributing to a complete census of the planetary population in the solar neighbourhood. The initial goals of the survey have therefore been fulfilled. The CARMENES sample continues to be observed within the CARMENES Legacy+ programme. The ultimate goal is to reach 50 measurements for all suitable targets (i.e. excluding multiples, RV-loud stars, etc.). The CARMENES Legacy+ extension of the survey is expected to run at least until the end of 2023 and, eventually, to lead to a second release of CARMENES survey data with 50 measurements or more for about 300 nearby M dwarfs. Through the present release and future additions, the CARMENES data will continue to yield new exoplanet discoveries and enable abundant studies in other domains within stellar astrophysics and exoplanetary science.

thumbnail Fig. 12

Comparison of low-mass/small planet occurrence rates from various surveys (adapted from Sab21). The parameter range is 1 to 10 M in Mpl sin i for RV surveys and 1.3 to 3.7 R in Rpl for transit surveys. The error bars for transit surveys are Poisson errors based on the number of planet detections in the respective bins, and error bars from RV surveys are the 16% and 84% levels of the occurrence rate posterior distributions. Results using transiting planets from the Kepler mission are represented as triangles (Morton & Swift 2014; Dressing & Charbonneau 2015; Mulders et al. 2015; Gaidos et al. 2016; Hsu et al. 2020), and rates from the HARPS, HADES, and CARMENES RV surveys are represented as squares (Bonfils et al. 2013; Sabotta et al. 2021; Pinamonti et al. 2022, and this work). The grey square shows the occurrence rate from this work with the assumption of a log-uniform distribution in Mpl sin i.

Acknowledgements

CARMENES is an instrument at the Centro Astronómico Hispano en Andalucía (CAHA) at Calar Alto (Almería, Spain), operated jointly by the Junta de Andalucía and the Instituto de Astrofísica de Andalucía (CSIC). CARMENES was funded by the Max-Planck-Gesellschaft (MPG), the Consejo Superior de Investigaciones Científicas (CSIC), the Ministerio de Economía y Competitividad (MINECO) and the European Regional Development Fund (ERDF) through projects FICTS-2011-02, ICTS-2017-07-CAHA-4, and CAHA16-CE-3978, and the members of the CARMENES Consortium (Max-Planck-Institut für Astronomie, Instituto de Astrofísica de Andalucía, Landessternwarte Königstuhl, Institut de Ciències de l’Espai, Institut für Astrophysik Göttingen, Universidad Complutense de Madrid, Thüringer Landessternwarte Tautenburg, Instituto de Astrofísica de Canarias, Hamburger Sternwarte, Centro de Astrobiología and Centro Astronómico Hispano-Alemán), with additional contributions by the MINECO, the Deutsche Forschungsgemeinschaft (DFG) through the Major Research Instrumentation Programme and Research Unit FOR2544 “Blue Planets around Red Stars”, the Klaus Tschira Stiftung, the states of Baden-Württemberg and Niedersachsen, and by the Junta de Andalucía. We acknowledge financial support from the Spanish Agencia Estatal de Investigación of the Ministerio de Ciencia e Innovación (AEI-MCIN) and the ERDF “A way of making Europe” through projects PID2020-117493GB-I00, PID2019-109522GB-C5[1:4], PID2019-110689RB-I00, PID2019-107061GB-C61, PID2019-107061GB-C64, PGC2018-098153-B-C33, PID2021-125627OB-C31/AEI/10.13039/501100011033, and the Centre of Excellence “Severo Ochoa” and “María de Maeztu” awards to the Institut de Ciències de l’Espai (CEX2020-001058-M), Instituto de Astrofísica de Canarias (CEX2019-000920-S), Instituto de Astrofísica de Andalucía (SEV-2017-0709), and Centro de Astrobiología (MDM-2017-0737). We also benefited from additional funding from: the Secretaria d’Universitats i Recerca del Departament d’Empresa i Coneixement de la Generalitat de Catalunya and the Agència de Gestió d’Ajuts Universitaris i de Recerca of the Generalitat de Catalunya, with additional funding from the European FEDER/ERDF funds, and from the Generalitat de Catalunya/CERCA programme; the DFG through the Major Research Instrumentation Programme and Research Unit FOR2544 “Blue Planets around Red Stars” (RE 2694/8-1); the University of La Laguna through the Margarita Salas Fellowship from the Spanish Ministerio de Universidades ref. UNI/551/2021-May-26, and under the EU Next Generation funds; the Gobierno de Canarias through projects ProID2021010128 and ProID2020010129; the Spanish MICINN under Ramón y Cajal programme RYC-2013-14875; the “Fondi di Ricerca Scientifica d’Ateneo 2021” of the University of Rome “Tor Vergata”; and the programme “Alien Earths” supported by the National Aeronautics and Space Administration (NASA) under agreement No. 80NSSC21K0593. This reasearch was based on data from the CARMENES data archive at CAB (CSIC-INTA), and made use of the NASA Exoplanet Archive, which is operated by the California Institute of Technology, under contract with the NASA under the Exoplanet Exploration Program. We thank B. Arroyo, A. Fernández, E. Gallego, A. Gardini, A. Guijarro, R. Hedrosa, I. Hermelo, J. Iglesias-Páramo, P. Martín, R. J. Mathar, M. Moreno, E. Ramos, R. Rebolo, I. Vico, and many other individuals for their commitment to the CARMENES GTO.

Appendix A Additional tables

Table A.1

Exoplanets in the CARMENES survey target sample with publications using CARMENES data.

Table A.2

Periodic signals with FAP < 1% in the 238-target sample for occurrence rate analysis.

References

  1. Abia, C., Tabernero, H. M., Korotin, S. A., et al. 2020, A&A, 642, A227 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  2. Affer, L., Micela, G., Damasso, M., et al. 2016, A&A, 593, A117 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  3. Alonso-Floriano, F. J., Morales, J. C., Caballero, J. A., et al. 2015, A&A, 577, A128 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  4. Amado, P. J., Bauer, F. F., Rodríguez López, C., et al. 2021, A&A, 650, A188 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  5. Anglada-Escudé, G., & Butler, R. P. 2012, ApJS, 200, 15 [Google Scholar]
  6. Anglada-Escudé, G., Boss, A. P., Weinberger, A. J., et al. 2012, ApJ, 746, 37 [Google Scholar]
  7. Astudillo-Defru, N., Díaz, R. F., Bonfils, X., et al. 2017a, A&A, 605, L11 [EDP Sciences] [Google Scholar]
  8. Astudillo-Defru, N., Forveille, T., Bonfils, X., et al. 2017b, A&A, 602, A88 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  9. Baranne, A., Queloz, D., Mayor, M., et al. 1996, A&AS, 119, 373 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  10. Baroch, D., Morales, J. C., Ribas, I., et al. 2018, A&A, 619, A32 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  11. Baroch, D., Morales, J. C., Ribas, I., et al. 2020, A&A, 641, A69 [EDP Sciences] [Google Scholar]
  12. Baroch, D., Morales, J. C., Ribas, I., et al. 2021, A&A, 653, A49 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  13. Bauer, F. F., Zechmeister, M., & Reiners, A. 2015, A&A, 581, A117 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  14. Bauer, F. F., Zechmeister, M., Kaminski, A., et al. 2020, A&A, 640, A50 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  15. Bayliss, D., Gillen, E., Eigmüller, P., et al. 2018, MNRAS, 475, 4467 [Google Scholar]
  16. Bitsch, B., Lambrechts, M., & Johansen, A. 2015, A&A, 582, A112 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  17. Blanco-Pozo, J., Perger, M., Damasso, M., et al. 2023, A&A, in press, https://doi.org/1S.1051/SSS4-6361/2S2245S53 [Google Scholar]
  18. Bluhm, P., Luque, R., Espinoza, N., et al. 2020, A&A, 639, A132 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  19. Bluhm, P., Pallé, E., Molaverdikhani, K., et al. 2021, A&A, 650, A78 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  20. Bonfils, X., Forveille, T., Delfosse, X., et al. 2005, A&A, 443, L15 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  21. Bonfils, X., Lo Curto, G., Correia, A. C. M., et al. 2013, A&A, 556, A110 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  22. Bonfils, X., Astudillo-Defru, N., Díaz, R., et al. 2018, A&A, 613, A25 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  23. Burn, R., Schlecker, M., Mordasini, C., et al. 2021, A&A, 656, A72 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  24. Burt, J., Vogt, S. S., Butler, R. P., et al. 2014, ApJ, 789, 114 [NASA ADS] [CrossRef] [Google Scholar]
  25. Butler, R. P., Johnson, J. A., Marcy, G. W., et al. 2006, PASP, 118, 1685 [Google Scholar]
  26. Butler, R. P., Howard, A. W., Vogt, S. S., & Wright, J. T. 2009, ApJ, 691, 1738 [NASA ADS] [CrossRef] [Google Scholar]
  27. Butler, R. P., Vogt, S. S., Laughlin, G., et al. 2017, AJ, 153, 208 [Google Scholar]
  28. Caballero, J. A., Guàrdia, J., López del Fresno, M., et al. 2016, in Proc. SPIE, 9910, 99100E [Google Scholar]
  29. Cadieux, C., Doyon, R., Plotnykov, M., et al. 2022, AJ, 164, 96 [NASA ADS] [CrossRef] [Google Scholar]
  30. Cale, B. L., Reefe, M., Plavchan, P., et al. 2021, AJ, 162, 295 [NASA ADS] [CrossRef] [Google Scholar]
  31. Carleo, I., Malavolta, L., Lanza, A. F., et al. 2020, A&A, 638, A5 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  32. Casasayas-Barris, N., Pallé, E., Yan, F., et al. 2020, A&A, 635, A206 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  33. Casasayas-Barris, N., Orell-Miquel, J., Stangret, M., et al. 2021, A&A, 654, A163 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  34. Chaturvedi, P., Bluhm, P., Nagel, E., et al. 2022, A&A, 666, A155 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  35. Cifuentes, C., Caballero, J. A., Cortés-Contreras, M., et al. 2020, A&A, 642, A115 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  36. Cloutier, R., Astudillo-Defru, N., Doyon, R., et al. 2017, A&A, 608, A35 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  37. Cortés-Contreras, M., Béjar, V. J. S., Caballero, J. A., et al. 2017, A&A, 597, A47 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  38. Czesla, S., Lampón, M., Sanz-Forcada, J., et al. 2022, A&A, 657, A6 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  39. Damasso, M., Perger, M., Almenara, J. M., et al. 2022, A&A, 666, A187 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  40. David, T. J., Hillenbrand, L. A., Petigura, E. A., et al. 2016, Nature, 534, 658 [CrossRef] [Google Scholar]
  41. Delfosse, X., Forveille, T., Beuzit, J. L., et al. 1999, A&A, 344, 897 [NASA ADS] [Google Scholar]
  42. Demory, B. O., Pozuelos, F. J., Gómez Maqueo Chew, Y., et al. 2020, A&A, 642, A49 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  43. Díaz, R. F., Delfosse, X., Hobson, M. J., et al. 2019, A&A, 625, A17 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  44. Díez Alonso, E., Caballero, J. A., Montes, D., et al. 2019, A&A, 621, A126 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  45. Dreizler, S., Crossfield, I. J. M., Kossakowski, D., et al. 2020, A&A, 644, A127 [EDP Sciences] [Google Scholar]
  46. Dressing, C. D., & Charbonneau, D. 2015, ApJ, 807, 45 [Google Scholar]
  47. Emsenhuber, A., Mordasini, C., Burn, R., et al. 2021, A&A, 656, A69 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  48. Endl, M., Cochran, W. D., Tull, R. G., & MacQueen, P. J. 2003, AJ, 126, 3099 [Google Scholar]
  49. Endl, M., Cochran, W. D., Wittenmyer, R. A., & Boss, A. P. 2008, ApJ, 673, 1165 [NASA ADS] [CrossRef] [Google Scholar]
  50. Espinoza, N., Pallé, E., Kemmer, J., et al. 2022, AJ, 163, 133 [NASA ADS] [CrossRef] [Google Scholar]
  51. Feng, F., Shectman, S. A., Clement, M. S., et al. 2020, ApJS, 250, 29 [Google Scholar]
  52. Forveille, T., Bonfils, X., Delfosse, X., et al. 2009, A&A, 493, 645 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  53. Fuhrmeister, B., Czesla, S., Schmitt, J. H. M. M., et al. 2018, A&A, 615, A14 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  54. Fuhrmeister, B., Czesla, S., Hildebrandt, L., et al. 2019a, A&A, 632, A24 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  55. Fuhrmeister, B., Czesla, S., Schmitt, J. H. M. M., et al. 2019b, A&A, 623, A24 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  56. Fuhrmeister, B., Czesla, S., Hildebrandt, L., et al. 2020, A&A, 640, A52 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  57. Fuhrmeister, B., Czesla, S., Nagel, E., et al. 2022, A&A, 657, A125 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  58. Gaia Collaboration (Smart, R.L., et al.) 2021, A&A, 649, A6 [Google Scholar]
  59. Gaidos, E., Mann, A. W., Kraus, A. L., & Ireland, M. 2016, MNRAS, 457, 2877 [Google Scholar]
  60. Garcia-Piquer, A., Morales, J. C., Ribas, I., et al. 2017, A&A, 604, A87 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  61. Ghezzi, L., Montet, B. T., & Johnson, J. A. 2018, ApJ, 860, 109 [Google Scholar]
  62. Gillon, M., Jehin, E., Lederer, S. M., et al. 2016, Nature, 533, 221 [Google Scholar]
  63. González-Álvarez, E., Zapatero Osorio, M. R., Caballero, J. A., et al. 2020, A&A, 637, A93 [EDP Sciences] [Google Scholar]
  64. González-Álvarez, E., Petralia, A., Micela, G., et al. 2021, A&A, 649, A157 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  65. González-Álvarez, E., Zapatero Osorio, M. R., Sanz-Forcada, J., et al. 2022, A&A, 658, A138 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  66. Hatzes, A. P. 2019, The Doppler Method for the Detection of Exoplanets (IOP Publishing) [Google Scholar]
  67. Hintz, D., Fuhrmeister, B., Czesla, S., et al. 2019, A&A, 623, A136 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  68. Hintz, D., Fuhrmeister, B., Czesla, S., et al. 2020, A&A, 638, A115 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  69. Hirano, T., Dai, F., Livingston, J. H., et al. 2018, AJ, 155, 124 [NASA ADS] [CrossRef] [Google Scholar]
  70. Hobson, M. J., Díaz, R. F., Delfosse, X., et al. 2018, A&A, 618, A103 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  71. Hobson, M. J., Delfosse, X., Astudillo-Defru, N., et al. 2019, A&A, 625, A18 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  72. Howard, A. W., Johnson, J. A., Marcy, G. W., et al. 2010, ApJ, 721, 1467 [Google Scholar]
  73. Howell, S. B., Sobeck, C., Haas, M., et al. 2014, PASP, 126, 398 [Google Scholar]
  74. Hoyer, S., Gandolfi, D., Armstrong, D. J., et al. 2021, MNRAS, 505, 3361 [NASA ADS] [CrossRef] [Google Scholar]
  75. Hsu, D. C., Ford, E. B., & Terrien, R. 2020, MNRAS, 498, 2249 [Google Scholar]
  76. Ida, S., & Lin, D. N. C. 2004, ApJ, 604, 388 [Google Scholar]
  77. Izidoro, A., Bitsch, B., Raymond, S. N., et al. 2021, A&A, 650, A152 [EDP Sciences] [Google Scholar]
  78. Jeffers, S. V., Schöfer, P., Lamert, A., et al. 2018, A&A, 614, A76 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  79. Jeffers, S. V., Barnes, J. R., Schöfer, P., et al. 2022, A&A, 663, A27 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  80. Johnson, J. A., Howard, A. W., Marcy, G. W., et al. 2010, PASP, 122, 149 [NASA ADS] [Google Scholar]
  81. Kaminski, A., Trifonov, T., Caballero, J. A., et al. 2018, A&A, 618, A115 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  82. Kanodia, S., & Wright, J. 2018, RNAAS, 2, 4 [NASA ADS] [Google Scholar]
  83. Kanodia, S., Wolfgang, A., Stefánsson, G. K., Ning, B., & Mahadevan, S. 2019, ApJ, 882, 38 [NASA ADS] [CrossRef] [Google Scholar]
  84. Kemmer, J., Stock, S., Kossakowski, D., et al. 2020, A&A, 642, A236 [EDP Sciences] [Google Scholar]
  85. Kemmer, J., Dreizler, S., Kossakowski, D., et al. 2022, A&A, 659, A17 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  86. Khalafinejad, S., Molaverdikhani, K., Blecic, J., et al. 2021, A&A, 656, A142 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  87. Kopparapu, R. K., Ramirez, R., Kasting, J. F., et al. 2013, ApJ, 765, 131 [NASA ADS] [CrossRef] [Google Scholar]
  88. Kossakowski, D., Kemmer, J., Bluhm, P., et al. 2021, A&A, 656, A124 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  89. Kossakowski, D., Kürster, M., Henning, T., et al. 2022, A&A, 666, A143 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  90. Kossakowski, D., Kürster, M., Trifonov, T., et al. 2023, A&A, 670, A84 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  91. Kürster, M., Endl, M., Rouesnel, F., et al. 2003, A&A, 403, 1077 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  92. Lafarga, M., Ribas, I., Lovis, C., et al. 2020, A&A, 636, A36 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  93. Lafarga, M., Ribas, I., Reiners, A., et al. 2021, A&A, 652, A28 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  94. Lalitha, S., Baroch, D., Morales, J. C., et al. 2019, A&A, 627, A116 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  95. Luque, R., & Pallé, E. 2022, Science, 377, 1211 [NASA ADS] [CrossRef] [Google Scholar]
  96. Luque, R., Nowak, G., Pallé, E., et al. 2018, A&A, 620, A171 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  97. Luque, R., Pallé, E., Kossakowski, D., et al. 2019, A&A, 628, A39 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  98. Luque, R., Fulton, B. J., Kunimoto, M., et al. 2022, A&A, 664, A199 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  99. Mahadevan, S., Stefánsson, G., Robertson, P., et al. 2021, ApJ, 919, L9 [NASA ADS] [CrossRef] [Google Scholar]
  100. Marfil, E., Tabernero, H. M., Montes, D., et al. 2020, MNRAS, 492, 5470 [NASA ADS] [CrossRef] [Google Scholar]
  101. Marfil, E., Tabernero, H. M., Montes, D., et al. 2021, A&A, 656, A162 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  102. Miguel, Y., Cridland, A., Ormel, C. W., Fortney, J. J., & Ida, S. 2020, MNRAS, 491, 1998 [NASA ADS] [Google Scholar]
  103. Mishra, L., Alibert, Y., Leleu, A., et al. 2021, A&A, 656, A74 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  104. Morales, J. C., Mustill, A. J., Ribas, I., et al. 2019, Science, 365, 1441 [Google Scholar]
  105. Mordasini, C., Alibert, Y., Klahr, H., & Henning, T. 2012, A&A, 547, A111 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  106. Morton, T. D., & Swift, J. 2014, ApJ, 791, 10 [Google Scholar]
  107. Mulders, G. D., Pascucci, I., & Apai, D. 2015, ApJ, 814, 130 [NASA ADS] [CrossRef] [Google Scholar]
  108. Nagel, E. 2019, PhD thesis, Universität Hamburg Germany, https://ediss.sub.uni-hamburg.de/handle/ediss/6280 [Google Scholar]
  109. Nagel, E., Czesla, S., Schmitt, J. H. M. M., et al. 2019, A&A, 622, A153 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  110. Nowak, G., Luque, R., Parviainen, H., et al. 2020, A&A, 642, A173 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  111. Nutzman, P., & Charbonneau, D. 2008, PASP, 120, 317 [Google Scholar]
  112. Oshagh, M., Bauer, F. F., Lafarga, M., et al. 2020, A&A, 643, A64 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  113. Pallé, E., Orell-Miquel, J., Brady, M., et al. 2023, A&A, submitted [arXiv:2301.06873] [Google Scholar]
  114. Passegger, V. M., Reiners, A., Jeffers, S. V., et al. 2018, A&A, 615, A6 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  115. Passegger, V. M., Schweitzer, A., Shulyak, D., et al. 2019, A&A, 627, A161 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  116. Passegger, V. M., Bello-García, A., Ordieres-Meré, J., et al. 2020, A&A, 642, A22 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  117. Passegger, V. M., Bello-García, A., Ordieres-Meré, J., et al. 2022, A&A, 658, A194 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  118. Perdelwitz, V., Mittag, M., Tal-Or, L., et al. 2021, A&A, 652, A116 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  119. Perger, M., García-Piquer, A., Ribas, I., et al. 2017, A&A, 598, A26 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  120. Perger, M., Scandariato, G., Ribas, I., et al. 2019, A&A, 624, A123 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  121. Perger, M., Ribas, I., Anglada-Escudé, G., et al. 2021, A&A, 649, L12 [EDP Sciences] [Google Scholar]
  122. Pinamonti, M., Sozzetti, A., Giacobbe, P., et al. 2019, A&A, 625, A126 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  123. Pinamonti, M., Sozzetti, A., Maldonado, J., et al. 2022, A&A, 664, A65 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  124. Piskunov, N. E., & Valenti, J. A. 2002, A&A, 385, 1095 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  125. Quirrenbach, A., Amado, P. J., Caballero, J. A., et al. 2014, in Proc. SPIE, 9147, 91471F [Google Scholar]
  126. Quirrenbach, A., Amado, P. J., Caballero, J. A., et al. 2016, in Proc. SPIE, 9908, 990812 [Google Scholar]
  127. Quirrenbach, A., Passegger, V. M., Trifonov, T., et al. 2022, A&A, 663, A48 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  128. Reiners, A., & Zechmeister, M. 2020, ApJS, 247, 11 [CrossRef] [Google Scholar]
  129. Reiners, A., Joshi, N., & Goldman, B. 2012, AJ, 143, 93 [Google Scholar]
  130. Reiners, A., Ribas, I., Zechmeister, M., et al. 2018a, A&A, 609, A5 [Google Scholar]
  131. Reiners, A., Zechmeister, M., Caballero, J. A., et al. 2018b, A&A, 612, A49 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  132. Reiners, A., Shulyak, D., Käpylä, P. J., et al. 2022, A&A, 662, A41 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  133. Reylé, C., Jardine, K., Fouqué, P., et al. 2021, A&A, 650, A201 [Google Scholar]
  134. Ribas, I., Tuomi, M., Reiners, A., et al. 2018, Nature, 563, 365 [Google Scholar]
  135. Ricker, G. R., Winn, J. N., Vanderspek, R., et al. 2015, J. Astron. Telescopes Instrum. Syst., 1, 014003 [Google Scholar]
  136. Robertson, P., Stefánsson, G., Mahadevan, S., et al. 2020, ApJ, 897, 125 [NASA ADS] [CrossRef] [Google Scholar]
  137. Sabotta, S., Schlecker, M., Chaturvedi, P., et al. 2021, A&A, 653, A114 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  138. Sánchez-López, A., López-Puertas, M., Snellen, I. A. G., et al. 2020, A&A, 643, A24 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  139. Sarkis, P., Henning, T., Kürster, M., et al. 2018, AJ, 155, 257 [NASA ADS] [CrossRef] [Google Scholar]
  140. Schlecker, M., Mordasini, C., Emsenhuber, A., et al. 2021a, A&A, 656, A71 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  141. Schlecker, M., Pham, D., Burn, R., et al. 2021b, A&A, 656, A73 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  142. Schlecker, M., Burn, R., Sabotta, S., et al. 2022, A&A, 664, A180 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  143. Schneider, J., Dedieu, C., Le Sidaner, P., Savalle, R., & Zolotukhin, I. 2011, A&A, 532, A79 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  144. Schöfer, P., Jeffers, S. V., Reiners, A., et al. 2019, A&A, 623, A44 [Google Scholar]
  145. Schweitzer, A., Passegger, V. M., Cifuentes, C., et al. 2019, A&A, 625, A68 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  146. Sedaghati, E., Sánchez-López, A., Czesla, S., et al. 2022, A&A, 659, A44 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  147. Seifahrt, A., Stürmer, J., Bean, J. L., & Schwab, C. 2018, SPIE Conf. Ser., 10702, 107026D [NASA ADS] [Google Scholar]
  148. Seifert, W., Sánchez Carrasco, M. A., Xu, W., et al. 2012, SPIE Conf. Ser., 8446, 844633 [NASA ADS] [Google Scholar]
  149. Shan, Y., Reiners, A., Fabbian, D., et al. 2021, A&A, 654, A118 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  150. Shulyak, D., Reiners, A., Nagel, E., et al. 2019, A&A, 626, A86 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  151. Soto, M. G., Anglada-Escudé, G., Dreizler, S., et al. 2021, A&A, 649, A144 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  152. Stefánsson, G., Cañas, C., Wisniewski, J., et al. 2020, AJ, 159, 100 [NASA ADS] [CrossRef] [Google Scholar]
  153. Stock, S., Kemmer, J., Reffert, S., et al. 2020a, A&A, 636, A119 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  154. Stock, S., Nagel, E., Kemmer, J., et al. 2020b, A&A, 643, A112 [EDP Sciences] [Google Scholar]
  155. Stürmer, J., Stahl, O., Schwab, C., et al. 2014, SPIE Conf. Ser., 9151, 915152 [Google Scholar]
  156. Suárez Mascareño, A., González Hernández, J. I., Rebolo, R., et al. 2017, A&A, 605, A92 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  157. Suárez Mascareño, A., Rebolo, R., González Hernández, J. I., et al. 2018, A&A, 612, A89 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  158. Suárez Mascareño, A., González-Álvarez, E., Zapatero Osorio, M. R., et al. 2023, A&A, 670, A5 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  159. Tal-Or, L., Zechmeister, M., Reiners, A., et al. 2018, A&A, 614, A122 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  160. Tal-Or, L., Trifonov, T., Zucker, S., Mazeh, T., & Zechmeister, M. 2019, MNRAS, 484, L8 [Google Scholar]
  161. Toledo-Padrón, B., Suárez Mascareño, A., González Hernández, J. I., et al. 2021, A&A, 648, A20 [EDP Sciences] [Google Scholar]
  162. Trifonov, T., Kürster, M., Zechmeister, M., et al. 2018, A&A, 609, A117 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  163. Trifonov, T., Lee, M. H., Kürster, M., et al. 2020a, A&A, 638, A16 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  164. Trifonov, T., Tal-Or, L., Zechmeister, M., et al. 2020b, A&A, 636, A74 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  165. Trifonov, T., Caballero, J. A., Morales, J. C., et al. 2021, Science, 371, 1038 [Google Scholar]
  166. Tuomi, M., Jones, H. R. A., Barnes, J. R., Anglada-Escudé, G., & Jenkins, J. S. 2014, MNRAS, 441, 1545 [Google Scholar]
  167. Tuomi, M., Jones, H. R. A., Barnes, J. R., et al. 2018, AJ, 155, 192 [NASA ADS] [CrossRef] [Google Scholar]
  168. Tuomi, M., Jones, H. R. A., Butler, R. P., et al. 2019, AAS J., submitted [arXiv: 1906.04644] [Google Scholar]
  169. Turtelboom, E. V., Weiss, L. M., Dressing, C. D., et al. 2022, AJ, 163, 293 [NASA ADS] [CrossRef] [Google Scholar]
  170. Wang, X.-Y., Rice, M., Wang, S., et al. 2022, ApJ, 926, L8 [NASA ADS] [CrossRef] [Google Scholar]
  171. Wright, J. T., & Eastman, J. D. 2014, PASP, 126, 838 [Google Scholar]
  172. Wright, J. T., Marcy, G. W., Butler, R. P., & Vogt, S. S. 2004, ApJS, 152, 261 [Google Scholar]
  173. Wright, D. J., Wittenmyer, R. A., Tinney, C. G., Bentley, J. S., & Zhao, J. 2016, ApJ, 817, L20 [NASA ADS] [CrossRef] [Google Scholar]
  174. Yan, F., Casasayas-Barris, N., Molaverdikhani, K., et al. 2019, A&A, 632, A69 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  175. Yan, F., Wyttenbach, A., Casasayas-Barris, N., et al. 2021, A&A, 645, A22 [EDP Sciences] [Google Scholar]
  176. Zechmeister, M., & Kürster, M. 2009, A&A, 496, 577 [CrossRef] [EDP Sciences] [Google Scholar]
  177. Zechmeister, M., Kürster, M., & Endl, M. 2009, A&A, 505, 859 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  178. Zechmeister, M., Anglada-Escudé, G., & Reiners, A. 2014, A&A, 561, A59 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  179. Zechmeister, M., Reiners, A., Amado, P. J., et al. 2018, A&A, 609, A12 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  180. Zechmeister, M., Dreizler, S., Ribas, I., et al. 2019, A&A, 627, A49 [NASA ADS] [EDP Sciences] [Google Scholar]

1

Calar Alto high-Resolution search for M dwarfs with Exoearths with Near-infrared and optical Échelle Spectrographs; https://carmenes.caha.es

3

CARMENES Reduction And CALibration.

4

SpEctrum Radial Velocity AnaLyser. https://github.com/mzechmeister/serval. Based on the version committed on 2022-01-26 https://github.com/mzechmeister/serval/tree/ a348b4c

5

Radial velocities and Activity indicators from Cross-COrrelatiON with masks. https://github.com/mlafarga/raccoon

9

The paper quotes 79 RVs, but only 72 RVs are published in https://cdsarc.cds.unistra.fr/ftp/J/A+A/618/A103/tablea1.dat

10

https://exoplanetarchive.ipac.caltech.edu, accessed on 1 July 2022.

12

The code and combined maps used to calculate occurrence rates are found in https://github.com/ssabotta/calculate_occurrence_rate

All Tables

Table 1

Basic properties and number of measurements for the CARMENES DR1 target sample.

Table 2

Nightly zero points for CARMENES.

Table 3

CARMENES survey targets with confirmed or claimed exoplanets that have no dedicated publications with CARMENES data.

Table 4

Exoplanets challenged by CARMENES.

Table 5

Planet occurrence rates for a sample of 238 M dwarfs in the CARMENES DR1 sample, including candidates and planet detections from other surveys.

Table A.1

Exoplanets in the CARMENES survey target sample with publications using CARMENES data.

Table A.2

Periodic signals with FAP < 1% in the 238-target sample for occurrence rate analysis.

All Figures

thumbnail Fig. 1

Distribution of the CARMENES GTO target sample (excluding the SB2 and ST3 systems) as a function of distance (d < 20 pc) for different spectral types or absolute Gaia G-band intervals. Some stars in the sample are at greater distances, and this number is provided inside the right-pointing arrow. One of the targets at the K–M spectral type boundary has an MG value below 7.73 mag and is not included; hence, the total number of stars plotted is 344. The distance distribution of the GCNS for the same intervals is also shown, and the ratios between the two are depicted as black crosses with the scale in the right y axis.

In the text
thumbnail Fig. 2

Nightly zero points (NZPs) of CARMENES VIS. The RVs of all RV constant stars (small green points) are zero-centred. A mean (the NZP, solid black circles) is computed for each night, in which identified outliers (small red points) are omitted. NZPs with fewer than three RV standard observations (open magenta circles) are replaced by a local mean.

In the text
thumbnail Fig. 3

Stacked histograms and statistical parameters (median and mode) of the internal precision (formal uncertainties) of the 18 642 precise RV measurements in the CARMENES DR1.

In the text
thumbnail Fig. 4

Distribution of observations. The rms is computed for each of the 344 targets (circles colour-coded by spectral type) from their serval RV time series (NZP-corrected). No periodic signals (activity or planetary) were removed. The faint target J16102−193 (K2–33) is excluded, since all its spectra were taken without F-P.

In the text
thumbnail Fig. 5

Time series data and periodograms for target J00051+457 (GJ 2). Top panel: NZP-corrected RV time series. Outlier measurements (grey symbols; see text) are excluded in the periodogram calculation. Bottom panels (from top to bottom): GLS periodograms of the NZP-corrected RVs (AVC, blue), the chromatic index (CRX, black) and CCF bisector span (BIS, red), the differential line width (dLW, black) and the CCF FWHM (red), and the chromospheric activity indices corresponding to the Ha (black) and averaged Ca II IRT lines (red). The dash-dotted blue vertical line in all panels marks the position of the most significant peak in the RV periodogram, while the circles in the sub-panels highlight the position of the strongest signal in each periodogram, with the period given in the legend. The 0.1%, 1%, and 10% FAP levels are shown as horizontal dashed, dash-dotted, and dotted green lines, respectively, and are visible if they fall within the displayed vertical range. Vertical axes in panels that show two different datasets are scaled such that the FAP levels are identical. The colour code allows the graph to be related with the axis labels and legends.

In the text
thumbnail Fig. 6

Same as Fig. 5, but for J02222+478 (GJ 96). No planetary signal at a period of 73.9 days, as claimed by Hobson et al. (2018), is visible in the CARMENES data. The only significant periodicity is at ~28.5 days and seems to be related to activity given the counterparts in some of the indicators. Phase-folded plots are shown in Fig. 7.

In the text
thumbnail Fig. 7

Phase-folded RV data of J02222+478 (GJ 96). Left. SOPHIE (green squares) and CARMENES VIS (blue circles) phase-folded to a period of 73.94 days. The planetary Keplerian signal (dashed green line) proposed by Hobson et al. (2018) is based on the SOPHIE data alone and disfavoured by the CARMENES measurements. Right same as the left panel, but phase-folded to the periodicity of 28.3 days, which is most likely associated with stellar activity.

In the text
thumbnail Fig. 8

Same as Fig. 5, but for J09561+627 (GJ 373). The periodic signal at 17.8 days attributed to a planet by Tuomi et al. (2019) and Feng et al. (2020) most likely arises from stellar activity in view of the counterparts in most of the activity indicators.

In the text
thumbnail Fig. 9

Scatter plots of the CARMENES DR1 exoplanet sample compared to the complete sample of catalogued planets in the NASA Exoplanet Archive detected via RVs (903; small dots). Different symbols indicate planets newly detected from the CARMENES blind survey (33; stars), planets confirmed from transit follow-up (26; circles), and known planets re-analysed with CARMENES data (17; triangles). The three panels correspond to pairs of different relevant parameters, with the complementary colour scale introducing a third dimension. The histograms along the axes show distributions of the corresponding parameters for the CARMENES planet sample. The blue shaded band in the top-right panel represents the liquid-water habitable zone with limits defined by the ‘runaway greenhouse’ and ‘maximum greenhouse’ criteria (Kopparapu et al. 2013).

In the text
thumbnail Fig. 10

CARMENES detection probability map in the period-mass plane derived from the observations of 238 targets. The detection probability is calculated using each target mass, and the average of all detection probabilities for the individual grid points is shown in the period-planet mass plane as a colour map. For example, a 10% detection probability for 238 targets indicates that the survey data can detect the respective planet in approximately 24 targets. Yellow circles indicate CARMENES planet detections and planet detections from other instruments that are confirmed by CARMENES data; grey circles planet candidates; and red circles planet detections from other surveys for which no sufficient CARMENES measurements are available to confirm them. The five planets in the low-probability detection region (<10%) are Teegarden’s Star b and c, YZ Cet c and d, and Wolf 1069 b, all of which are Earth-mass planets with host stars of less than 0.2 M, thus suggesting that such planets are very abundant.

In the text
thumbnail Fig. 11

Histogram of the number of planets detected in six Mpl sin i bins. The squares indicate the number of planets corrected for the survey sensitivity averaged over periods of up to 240 days. The dashed line and the solid line indicate the power law fit for Mpl < 32 M and Mpl > 32 M. respectively.

In the text
thumbnail Fig. 12

Comparison of low-mass/small planet occurrence rates from various surveys (adapted from Sab21). The parameter range is 1 to 10 M in Mpl sin i for RV surveys and 1.3 to 3.7 R in Rpl for transit surveys. The error bars for transit surveys are Poisson errors based on the number of planet detections in the respective bins, and error bars from RV surveys are the 16% and 84% levels of the occurrence rate posterior distributions. Results using transiting planets from the Kepler mission are represented as triangles (Morton & Swift 2014; Dressing & Charbonneau 2015; Mulders et al. 2015; Gaidos et al. 2016; Hsu et al. 2020), and rates from the HARPS, HADES, and CARMENES RV surveys are represented as squares (Bonfils et al. 2013; Sabotta et al. 2021; Pinamonti et al. 2022, and this work). The grey square shows the occurrence rate from this work with the assumption of a log-uniform distribution in Mpl sin i.

In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.