Using birefringent elements and imaging Michelsons for calibration of high precision planet finding spectrographs

TL;DR

This paper proposes a novel calibration method for high-precision spectrographs using birefringent elements and imaging Michelsons, offering a compact, stable, and full-spectrum alternative to iodine cells for exoplanet detection.

Contribution

It introduces a new calibration technique employing birefringent elements and Michelson interferometers, improving coverage, stability, and compactness over existing iodine cell methods.

Findings

Potential calibration stability at 0.1 m/s over relevant wavelengths.

Design can be made compact and stable for spectrograph calibration.

Method covers the entire visible spectrum with temperature drift control.

Abstract

One of the main methods used for finding extrasolar planets is the radial velocity technique, in which the Doppler shift of a star due to an orbiting planet is measured. These measurements are typically performed using cross-dispersed echelle spectrographs. Unfortunately such spectrographs are large and expensive and their accurate calibration continues to be challenging. The aim is to develop a different way to provide a calibration signal. A commonly used way to introduce a calibration signal is to insert an iodine cell in the beam. Disadvantages of this include that the lines are narrow, do not cover the entire spectrum and that light is absorbed. Here I show that inserting a birefringent element or an imaging Michelson, combined with Wollaston prisms eliminates these three shortcomings, while maintaining most of the benefits of the iodine approach. The proposed designs can be made very compact, thereby providing a convenient way of calibrating a spectrograph. Similar to the iodine cell approach, the calibration signal travels with the stellar signal, thereby reducing the sensitivity to spectrograph stability. The imposed signal covers the entire visible range and any temperature drifts will be consistent and describable by a single number. Based on experience with similar devices used, in a different configuration, by the Helioseismic and Magnetic Imager, it is shown that the calibration device can be made stable at the 0.1 m/s level, over a significant wavelength range, on short to medium time scales. While promising, many details still need to be worked out. In particular a number of laboratory measurements are required in order to finalize a design and estimate actual performance and it would be desirable to make a proof of concept.