A Quiet Error in Planet Hunting express
In a quiet observatory, a spectrograph catches the light of a distant star. The instrument spreads it into a rainbow — the spectrum. Lines in the spectrum shift slightly back and forth: an invisible planet is tugging the star. To spot this shift, astronomers compare the spectrum with a reference — a "portrait" of the star assembled from many observations.
One comparison method is the derivative method. It looks at how quickly the brightness changes along the spectrum and uses that slope to compute the shift. But if the reference portrait is built from faint frames (with a low signal-to-noise ratio), the slope becomes inaccurate. A translator is distracted by every rustle in the hall, and a one-hour meeting stretches into two. The authors tested this on artificial lines and real data.
When the signal becomes brighter (signal-to-noise above 1000), the error drops below the sensitivity threshold of the best modern spectrographs. And if you average many lines, the uncertainty shrinks at any quality. On real observations, the difference between methods was only 12.5 centimeters per second for a template made from 78 frames — less than the growth rate of bamboo.
This work is a reminder: in the hunt for Earth-sized planets, even a microscopic glitch in calculations can mislead. A quiet error in the method, almost imperceptible, can mask a real signal. And then the spectrograph in the quiet observatory will keep catching light, unaware of the hidden planet.
Original abstract
as written by the authors arXiv:2606.30377
The radial velocity (RV) method is a widely used technique to infer planetary masses and orbital parameters. One of the most widely used methods to compute RVs is based on the alignment of a high-SNR, data driven, stellar model with individual observations, commonly referred to as template matching. Typically, the alignment is performed through a $$\chi^2$$ minimization, but some approaches rely on the derivative of the stellar template to do so, which is often the case in line-by-line methods. In this paper we aim to explore the limitation of derivative-based methods for RV extraction in the case of using low-SNR stellar models. We use simulated Gaussian profiles to investigate the effect of computing a numerical derivative of the stellar template, in comparison with the usage of an analytical profile. The impact on RV and associated uncertainty is then analyzed as a function of the signal-to-noise ratio (SNR) of the line. Then, using real observations we compare the residuals between a derivative-based RV extraction and a classical template matching implementation, as function of the SNR of the stellar template. We find that on simulated Gaussian profiles the usage of the numerical approach leads to a RV residual at the level of the meter per second, at a per-pixel SNR regime of 100. An increase of the SNR leads to a decrease of this impact, falling under the current noise-floor of state-of-the-art spectrographs at SNR>1000 for a single spectral line. The inclusion of multiple spectral lines in the simulations lead to an overall decrease of the contamination, across all SNR regimes. The application to a real dataset presents a decrease of the RV impact with the increase of the template's SNR, albeit still presenting a 12.5 cm/s difference when including 78 observations in the stellar template.
The radial velocity method first helped find an exoplanet around the star 51 Pegasi in 1995. Since then, precision has improved so much that we can detect a star's wobble of centimeters per second — slower than a snail's pace.