# The Measured Impact of Chromatic Atmospheric Effects on Barycentric Corrections: Results from the EXtreme PREcision Spectrograph

## TL;DR

This study demonstrates that chromatic atmospheric effects significantly impact radial velocity measurements in exoplanet detection, and implementing wavelength-dependent barycentric corrections with the EXPRES instrument reduces errors below the noise floor.

## Contribution

The paper introduces a system for chromatic photon-weighted barycentric corrections and provides empirical results from its first year of operation with EXPRES.

## Key findings

- Ignoring chromatic corrections can cause up to 1 m/s radial velocity errors.
- Frequent flux sampling reduces errors to below the instrumental noise floor.
- Errors vary with stellar type and atmospheric conditions.

## Abstract

One source of error in high-precision radial velocity measurements of exoplanet host stars is chromatic change in Earth's atmospheric transmission during observations. Mitigation of this error requires that the photon-weighted barycentric correction be applied as a function of wavelength across the stellar spectrum. We have designed a system for chromatic photon-weighted barycentric corrections with the EXtreme PREcision Spectrograph (EXPRES) and present results from the first year of operations, based on radial velocity measurements of more than $10^3$ high-resolution stellar spectra. For observation times longer than 250 seconds, we find that if the chromatic component of the barycentric corrections is ignored, a range of radial velocity errors up to 1 m s$^{-1}$ can be incurred with cross-correlation, depending on the nightly atmospheric conditions. For this distribution of errors, the standard deviation is 8.4 cm s$^{-1}$ for G-type stars, 8.5 cm s$^{-1}$ for K-type stars, and 2.1 cm s$^{-1}$ for M-type stars. This error is reduced to well-below the instrumental and photon-noise limited floor by frequent flux sampling of the observed star with a low-resolution exposure meter spectrograph.