A systematic approach to determining the properties of an iodine absorption cell for high-precision radial velocity measurements

TL;DR

This paper presents a new method for measuring the effective temperature inside iodine absorption cells used for high-precision radial velocity measurements, analyzing the effects of temperature on wavelength stability and velocity shifts.

Contribution

A novel approach utilizing Boltzmann distribution of line depths to determine the molecular temperature inside iodine cells, improving understanding of temperature effects on radial velocity precision.

Findings

Radial velocity precision is independent of cell temperature.

Wavelength dependency is influenced by spectrometer resolving power and SNR.

Temperature crossing the dew point causes abrupt velocity jumps of 50 m/s.

Abstract

Absorption cells filled with diatomic iodine are frequently employed as wavelength reference for high-precision stellar radial velocity determination due their long-term stability and low cost. Despite their wide-spread usage in the community, there is little documentation on how to determine the ideal operating temperature of an individual cell. We have developed a new approach to measuring the effective molecular temperature inside a gas absorption cell and searching for effects detrimental to a high precision wavelength reference, utilizing the Boltzmann distribution of relative line depths within absorption bands of single vibrational transitions. With a high resolution Fourier transform spectrometer, we took a series of 632 spectra at temperatures between 23{\deg}C and 66{\deg}C. These spectra provide a sufficient basis to test the algorithm and demonstrate the stability and repeatability of the temperature determination via molecular lines on a single iodine absorption cell. The achievable radial velocity precision is found to be independent of the cell temperature and a detailed analysis shows a wavelength dependency, which originates in the resolving power of the spectrometer in use and the signal-to-noise ratio. Two effects were found to cause apparent absolute shifts in radial velocity, a temperature-induced shift of the order of 1 m/s/K and a more significant effect resulting in abrupt jumps of 50 m/s is determined to be caused by the temperature crossing the dew point of the molecular iodine.