J/A+A/687/A148 HD 158259 SOPHIE correction of RV systematics (Grouffal+, 2024)
An improved correction of radial velocity systematics for the
SOPHIE spectrograph.
Grouffal S., Santerne A., Hara N.C., Boisse I., Coez S., Heidari N.,
Sulis S.
<Astron. Astrophys. 687, A148 (2024)>
=2024A&A...687A.148G 2024A&A...687A.148G (SIMBAD/NED BibCode)
ADC_Keywords: Stars, bright ; Radial velocities ; Optical
Keywords: instrumentation: spectrographs - methods: data analysis -
techniques: radial velocities - planetary systems -
stars: individual: HD 158259
Abstract:
High-precision spectrographs can on occasion exhibit temporal
variations in their reference velocity or nightly zero point (NZP).
One way to monitor the NZP is to measure bright stars, whose intrinsic
radial velocity variation is assumed to be much smaller than the
instrument precision. The variations of these bright stars, which is
primarily assumed to be instrumental, are then smoothed into a
reference radial velocity time series (master constant) that is
subtracted from the observed targets. While this method is effective
in most cases, it does not fully propagate the uncertainty arising
from NZP variations. We present a new method for correcting for NZP
variations in radial velocity time series. This method uses Gaussian
processes based on ancillary information to model these systematic
effects. Moreover, it enables us to propagate the uncertainties of
this correction into the overall error budget. Another advantage of
this approach is that it relies on ancillary data that are collected
simultaneously with the spectra and does not solely depend on
dedicated observations of constant stars. We applied this method to
the SOPHIE spectrograph at the Haute-Provence Observatory using a few
instrument housekeeping data, such as the internal pressure and
temperature variations. Our results demonstrate that this method
effectively models the red noise of constant stars, even with a
limited number of housekeeping data, while preserving the signals of
exoplanets. Using simulations with mock planets and real data, we
found that this method significantly improves the false-alarm
probability of detections. It improves the probability by several
orders of magnitude. Additionally, by simulating numerous planetary
signals, we were able to detect up to 10% more planets with
small-amplitude radial velocity signals. We used this new correction
to reanalyse the planetary system around HD158259 and to improve the
detection of the outermost planets. We propose this technique as a
complementary approach to the classical master-constant correction of
the instrumental red noise. We also suggest to decrease the observing
cadence of the constant stars to optimise the telescope time for
scientific targets.
Description:
We present an improved method to correct for instrumental variations
in SOPHIE (OHP) data. The original method was based on the observation
of constant stars and the removing of the sliding median of the
constant star from other stars. This new method uses housekeeping
variables from the instrument (pressure and temperature) and Gaussian
Processes to correct for night-to-night variations.
We present the table of pressure and temperatures measurement from the
SOPHIE spectrograph from 2012 to 2022 and the result of the Gaussian
Processes on constant stars.
The authors would like to thank the team at OHP and in particular
Francois Moreau for the pressure and temperature measurements.
Objects:
----------------------------------------------------
RA (2000) DE Designation(s)
----------------------------------------------------
17 25 24.05 +52 47 26.4 HD158259 = TOI-1462
----------------------------------------------------
File Summary:
--------------------------------------------------------------------------------
FileName Lrecl Records Explanations
--------------------------------------------------------------------------------
ReadMe 80 . This file
tabled1.dat 75 2953 Result of the Gaussian Process training on
constant stars
tablef1.dat 344 845460 Temperature and pressure measurements from the
SOPHIE spectrograph from 2012 to 2022
--------------------------------------------------------------------------------
See also:
J/A+A/636/L6 : HD 158259 SOPHIE radial velocities (Hara+, 2020)
Byte-by-byte Description of file: tabled1.dat
--------------------------------------------------------------------------------
Bytes Format Units Label Explanations
--------------------------------------------------------------------------------
1- 24 E24.18 d Date Date (BJD)
26- 50 E25.18 m/s RV GP constant star value
52- 75 E24.18 m/s e_RV Error on the GP constant star
--------------------------------------------------------------------------------
Byte-by-byte Description of file: tablef1.dat
--------------------------------------------------------------------------------
Bytes Format Units Label Explanations
--------------------------------------------------------------------------------
1- 19 A19 "datime" Obs.date Date of the measurement
21- 44 E24.18 K Tcontainer Temperature of the air in the container
46- 69 E24.18 K Tgrating Temperature of the grating
71- 94 E24.18 K TbenchW Temperature of the bench West
96-119 E24.18 K TbenchE Temperature of the bench East
121-144 E24.18 K Tcryostat Temperature of at the cryostat
146-169 E24.18 K Tcoude Temperature at the Coude tube
171-194 E24.18 K Tferrule Temperature of the ferrule
196-219 E24.18 K Tshutter Temperature of the shutter
221-244 E24.18 K Tspectro Temperature of the air in the insulated
envelope (spectrograph)
246-269 E24.18 K Tobsroom Temperature in the observation room
271-294 E24.18 K Telecroom Temperature in the electronic room
296-319 E24.18 Pa Patmo Atmospheric pressure
321-344 E24.18 Pa Ptank Pressure in the tank
--------------------------------------------------------------------------------
Acknowledgements:
Salome Grouffal, salome.grouffal(at)lam.fr
(End) Patricia Vannier [CDS] 17-May-2024