The EXtreme PREcision Spectrograph (EXPRES) is an environmentally stabilized, fiber-fed, R = 137, 500, optical spectrograph. It was recently commissioned at the 4.3-m Lowell Discovery Telescope (LDT) near Flagstaff, Arizona. The spectrograph was designed with a target radial-velocity (RV) precision of 30 cm s −1 . In addition to instrumental innovations, the EXPRES pipeline, presented here, is the first for an on-sky, optical, fiber-fed spectrograph to employ many novel techniques-including an "extended flat" fiber used for wavelengthdependent quantum efficiency characterization of the CCD, a flat-relative optimal extraction algorithm, chromatic barycentric corrections, chromatic calibration offsets, and an ultra-precise laser frequency comb for wavelength calibration. We describe the reduction, calibration, and radial-velocity analysis pipeline used for EXPRES and present an example of our current sub-meter-per-second RV measurement precision, which reaches a formal, single-measurement error of 0.3 m s −1 for an observation with a per-pixel signal-to-noise ratio of 250. These velocities yield an orbital solution on the known exoplanet host 51 Peg that matches literature values with a residual RMS of 0.895 m s −1 .

The EXtreme PREcision Spectrograph (EXPRES) is a new Doppler spectrograph designed to reach a radialvelocity measurement precision sufficient to detect Earth-like exoplanets orbiting nearby, bright stars. We report on extensive laboratory testing and on-sky observations to quantitatively assess the instrumental radial-velocity measurement precision of EXPRES, with a focused discussion of individual terms in the instrument error budget. We find that EXPRES can reach a single-measurement instrument calibration precision better than 10 cm s −1 , not including photon noise from stellar observations. We also report on the performance of the various environmental, mechanical, and optical subsystems of EXPRES, assessing any contributions to radial-velocity error. For atmospheric and telescope related effects, this includes the fast tip-tilt guiding system, atmospheric dispersion compensation, and the chromatic exposure meter. For instrument calibration, this includes the laser fRequency comb (LFC), flat-field light source, CCD detector, and effects in the optical fibers. Modal noise is mitigated to a negligible level via a chaotic fiber agitator, which is especially important for wavelength calibration with the LFC. Regarding detector effects, we empirically assess the impact on the radial-velocity precision due to pixel-position nonuniformities and charge transfer inefficiency (CTI). EXPRES has begun its science survey to discover exoplanets orbiting G-dwarf and K-dwarf stars, in addition to transit spectroscopy and measurements of the Rossiter-McLaughlin effect.

The next generation of exoplanet-hunting spectrographs should deliver up to an order of magnitude improvement in radial velocity (RV) precision over the standard 1 state-of-the-art spectrographs. This advance is critical for enabling the detection of Earth-mass planets around Sun-like stars. New calibration techniques such as laser frequency combs and stabilized etalons ensure that the instrumental stability is well characterized. However, additional sources of error include stellar noise, undetected short-period planets, and telluric contamination. To understand and ultimately mitigate error sources, the contributing terms in the error budget must be isolated to the greatest extent possible. Here, we introduce a new high-cadence RV program, the Extreme Precision Spectrograph (EXPRES) 100  Earths  Survey, which aims to identify rocky planets around bright, nearby G and K dwarfs. We also present a benchmark case: the 62 day orbit of a Saturn-mass planet orbiting the chromospherically quiet star, HD 3651. The combination of high eccentricity (0.6) and a moderately long orbital period ensures significant dynamical clearing of any inner planets. Our Keplerian model for this planetary orbit has a residual rms of 58 cm s−1 over a ∼6 month time baseline. By eliminating significant contributors to the RV error budget, HD 3651 serves as a standard for evaluating the long-term precision of extreme precision RV programs.

Optical fiber modal noise is a limiting factor for high precision spectroscopy signal-to-noise in the near-infrared and visible. Unabated, especially when using highly coherent light sources for wavelength calibration, modal noise can induce radial velocity (RV) errors that hinder the discovery of low-mass (and potentially Earth-like) planets. Previous research in this field has found sufficient modal noise mitigation through the use of an integrating sphere, but this requires extremely bright light sources, a luxury not necessarily afforded by the next generation of high-resolution optical spectrographs. Otherwise, mechanical agitation, which "mixes" the fiber's modal patterns and allows the noise to be averaged over minutes-long exposures, provides some noise reduction but the exact mechanism behind improvement in signal-to-noise and RV drift has not been fully explored or optimized by the community. Therefore, we have filled out the parameter space of modal noise agitation techniques in order to better understand agitation's contribution to mitigating modal noise and to discover a better method for agitating fibers. We find that modal noise is best suppressed by the quasi-chaotic motion of two high-amplitude agitators oscillating with varying phase for fibers with large core diameters and low azimuthal symmetry. This work has subsequently influenced the design of a fiber agitator, to be installed with the EXtreme PREcision Spectrograph, that we estimate will reduce modal-noise-induced RV error to less than 3.2 cm s −1 . arXiv:1802.01642v1 [astro-ph.IM] 5 Feb 2018 Systematic variationsSince the resultant speckle pattern is dependent on dynamic optical properties of the fiber, modal noise can induce false RV's on the spectrograph (Mahadevan et al.

By controlling instrumental errors to below 10 cm s−1, the EXtreme PREcision Spectrograph (EXPRES) allows for a more insightful study of photospheric velocities that can mask weak Keplerian signals. Gaussian processes (GP) have become a standard tool for modeling correlated noise in radial velocity data sets. While GPs are constrained and motivated by physical properties of the star, in some cases they are still flexible enough to absorb unresolved Keplerian signals. We apply GP regression to EXPRES radial velocity measurements of the 3.5 Gyr old chromospherically active Sun-like star, HD 101501. We obtain tight constraints on the stellar rotation period and the evolution of spot distributions using 28 seasons of ground-based photometry, as well as recent Transiting Exoplanet Survey Satellite data. Light-curve inversion was carried out on both photometry data sets to reveal the spot distribution and spot evolution timescales on the star. We find that the >5 m s−1 rms radial velocity variations in HD 101501 are well modeled with a GP stellar activity model without planets, yielding a residual rms scatter of 45 cm s−1. We carry out simulations, injecting and recovering signals with the GP framework, to demonstrate that high-cadence observations are required to use GPs most efficiently to detect low-mass planets around active stars like HD 101501. Sparse sampling prevents GPs from learning the correlated noise structure and can allow it to absorb prospective Keplerian signals. We quantify the moderate to high-cadence monitoring that provides the necessary information to disentangle photospheric features using GPs and to detect planets around active stars.