a Radial Velocity Reference Jeff Valenti (STScI) Jay Anderson - - PowerPoint PPT Presentation

a radial velocity reference
SMART_READER_LITE
LIVE PREVIEW

a Radial Velocity Reference Jeff Valenti (STScI) Jay Anderson - - PowerPoint PPT Presentation

Using a Gas Absorption Cell as a Radial Velocity Reference Jeff Valenti (STScI) Jay Anderson (STScI) Presented at Astronomy of Exoplanets with Precise Radial Velocities at Penn State University on Aug 18, 2010 Why a gas cell can be


slide-1
SLIDE 1

“Using a Gas Absorption Cell as a Radial Velocity Reference” Jeff Valenti (STScI) Jay Anderson (STScI)

Presented at “Astronomy of Exoplanets with Precise Radial Velocities” at Penn State University on Aug 18, 2010

slide-2
SLIDE 2

Why a gas cell can be useful… A gas cell imprints on each spectrum the behavior of optics and detector for the actual illumination conditions during that observation Compensate for spectrograph instabilities. Data analysis is nontrivial. Planets still lurking in 15 years of existing data from slit spectrographs.

slide-3
SLIDE 3

Outline Modeling observations Intrinsic stellar spectrum Iodine cell temperature Line spread function Residuals Results

slide-4
SLIDE 4

Using a Gas Absorption Cell Model calculation

Determine wavelength scale of observation Shift intrinsic stellar spectrum by stellar radial velocity Multiply by gas cell transmission spectrum Convolve with local line spread function Determine normalization function to match observation

Free parameters for each observation

Wavelength scale Stellar radial velocity Normalization function Line spread function

slide-5
SLIDE 5

Wavelengths from Iodine Cell Absorption Lines

slide-6
SLIDE 6

Velocity Shift of Intrinsic Stellar Sepctrum

slide-7
SLIDE 7

Line Spread Function of Spectrograph

slide-8
SLIDE 8

Constructed Model of Observation

slide-9
SLIDE 9

Outline Modeling observations Intrinsic stellar spectrum Iodine cell temperature Line spread function Residuals Results

slide-10
SLIDE 10

Three Ways to Determine the Intrinsic Spectrum Observe directly with R ~ 300 000 spectrograph Deconvolve using contemporaneous LSF

Observe B stars with iodine to get an LSF Observe target star without iodine Deconvolve to get intrinsic stellar spectrum Assumes LSF is stable between observations

Deconvolve using simultaneous LSF

Observe target star several/many times with iodine “Grand solution” gives LSF and intrinsic stellar spectrum Still working to understand and tune the algorithm

slide-11
SLIDE 11

Deconvolution using Contemporaneous LSF

slide-12
SLIDE 12

Plenty of Constraints for Grand Solution

New Code

slide-13
SLIDE 13

Stellar Spectrum Rings if Nodes Too Far Apart

New Code

Same Set of Reduced Spectra Completely New Analysis Code

slide-14
SLIDE 14

Stellar Spectra Deconvolved Two Different Ways

slide-15
SLIDE 15

Outline Modeling observations Intrinsic stellar spectrum Iodine cell temperature Line spread function Residuals Results

slide-16
SLIDE 16

Transmission Spectrum of Keck Iodine Cell FTS spectra at three iodine cell temperatures

50, 55, and 60 C Interpolate to other temperatures as needed

slide-17
SLIDE 17

Temperature Sensitivity of Iodine Lines

slide-18
SLIDE 18

Iodine Cell Temperature vs. TEMPIOD1

Temperature variation, but velocities are good

slide-19
SLIDE 19

Environment Can Affect Gas Cell Temperature

TIOD1 TIOD2 TIN Radiative Cooling Control Sensor Second Sensor Thermal Control Stabilized Sensor Second Sensor Calibration Mirror In Out Environment

slide-20
SLIDE 20

Outline Modeling observations Intrinsic stellar spectrum Iodine cell temperature Line spread function Residuals Results

slide-21
SLIDE 21

LSF Changes For Each Exposure Consecutive exposures

67 second cadence

Raw LSF shift

0.0039 pixels 5.2 m/s

After modeling I2

0.5 m/s Factor of 10 better

slide-22
SLIDE 22

LSF Variations for Consecutive Exposures Spectrograph is stable on short time scales Slit illumination may vary

Misguiding Seeing changes

Pupil illumination may vary

Misguiding with telescope out of focus Particular concern for mosaic gratings

Reduce effects with spectrograph design

Fiber feed Precise guiding

slide-23
SLIDE 23

Spline Nodes Describe Narrow LSF Core

Free Centroid at Zero Fixed

New Code

slide-24
SLIDE 24

Works Equally Well for Broader LSF Core

New Code

slide-25
SLIDE 25

Broad LSF Wings Seen in Laser Exposures

5939.32 Å 5433.65 Å 6328.16 Å 1.3% of LSF is

  • utside ±7 pixels
slide-26
SLIDE 26

Outline Modeling observations Intrinsic stellar spectrum Iodine cell temperature Line spread function Residuals Results

slide-27
SLIDE 27

Fit Residuals for B Star Spectra

New Code

slide-28
SLIDE 28

Fit Residuals for 992 B Star Spectra

New Code

slide-29
SLIDE 29

Adjusted Fit Residuals for 992 B Stars

New Code Systematics reduced but not yet eliminated

slide-30
SLIDE 30

σ Dra without Residual Correction

New Code Prior to Reducing Fit Residuals

slide-31
SLIDE 31

σ Dra with Residual Correction and Uniform BC

New Code After Reducing Fit Residuals Systematics reduced but not yet eliminated

slide-32
SLIDE 32

Outline Modeling observations Intrinsic stellar spectrum Iodine cell temperature Line spread function Residuals Results

slide-33
SLIDE 33

Radial Velocities for τ Cet

New Code

slide-34
SLIDE 34

Radial Velocities for HD 9407

New Code

slide-35
SLIDE 35

Radial Velocities for HD 156668

New Code

slide-36
SLIDE 36

Radial Velocities for GJ 412a

1 pixel per node in intrinsic spectrum New Code

slide-37
SLIDE 37

Main Points

Gas cell compensates for spectrograph instabilities Need Instrinsic stellar spectrum

Obtain directly with R ~ 300 000 spectrograph Deconvolve using contemporaneous LSF Deconvolve using simultaneous LSF (“grand solution”)

Iodine cell temperature depends on environment Describe LSF by spline curve

Centroid at zero breaks degeneracy with wavelengths Need to accommodate extended wings seen in laser

Diagnostics of systematic errors

Fit residuals of many stars in iodine reference frame Radial velocity versus barycentric correction

Grand solution is starting to yield precise velocities