HST/WFC3 Flux Calibration Ladder: Vega
SPACE
TELESCOPE
SCIENCE
INSTITUTE
Susana E. Deustua,, Ralph Bohlin, Nor Pirzkal & John MaKenty. Space Telescope Science Institute (United States)
Operated for NASA by AURA
Vega Infrared Spectra!
Introduction!
G141 ibtw01
!
!
!
Vega is the quintessential absolute flux calibrator in Astronomy, and,
!
one of only a few stars calibrated against an SI-traceable blackbody. The
!
majority of experiments made measurements the visible (see Megessier
Vega: coadded spectra G141
!
8•10
1995 for a detailed discussion) with uncertainties on the order of a few
!
percent. Equivalent measurements made in the infrared, from the
6•10
!
ground, yielded much larger uncertainties (e.g. Blackwell and
4•10
Figure 3. Visit 01; slow scan, -1st order spectra in
collaborators), due in large part to atmospheric effects.
Hence,
G141 coadded at each scan position, after dark
Figure 1. Positions on the detector of the spectral scans. Grism orders
accurate measurements made above the atmosphere should reduce the
2•10 right to left are +2nd, +1st, 0th, -1st and -2nd. Left: 0th, -1st, -2nd.
subtraction, flatfielding, dispersion correction and
from
rectification. Note the 10% variation in the counts
Slivers of the +1st on the right edge, and -3rd on the left edge are visible.
uncertainties!
rates at wavelengths longer than 1.2 microns (left
Center:
+1st, 0th, -1st and -2nd orders. Right: +2nd, +1st, 0th and -1st
0
1.0 visible.
1.2 The1.4
1.6 and1.8
side of the plot). This is likely due to flat field
orders are
0th, +1st
+2nd orders are saturated!
!
Wavelength (microns)
variations.!
STIS spectral observations of Vega between 3000 angstroms and 1
Vega: coadded spectra G141
Vega: coadded spectra G102
8•10
6•10
micron were made by Bohlin and Gilliland (2004, AJ, 127, 3508 ) in the
traditional stare mode. By exploiting the fact that in the STIS CCD
6•10
4•10
saturated charge in one pixel spills over to neighboring pixels, thereby
4•10
conserving flux, they were able to obtain high signal to noise spectra of
2•10
2•10
this fundamental flux standard. !
!
0
0
1.0
1.2
1.4
1.6
1.8
0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
At
wavelengths longer than 1 micron Vega’s spectral energy
Wavelength (microns)
Wavelength (microns)
distribution is obtained by extrapolating from the current UVIS data via
coadded
spectra
G102
Figure 2 Coadded unfluxed spectra of Vega Vega:
at each
scan rate
for the
-1st orders
6•10
models. To fill in the crucial gap between 0.9 and 1.7 microns, we
of the G141 (top) and G102 (bottom) WFC3 IR grisms. Wavelength scale is in
microns, negative numbers indicate negative order. Solid lines are the slow scan
started a program to acquire grism spectroscopy of Vega in the near
Figure 4. Comparison of GD71 and Vega
rates; dashed lines are the fast scans.
Top:
Paschen
β
is
the
dip
at
-1.28
4•10
spectra showing the wavelength calibration of
microns, and the series of features between 1.6 and 1.7 microns are Br 13,12,
infrared using the two Wide Field Camera 3 (WFC3) infrared grisms, in
the -1st order is comparable to the +1st order
11. Bottom: Pa γ is just visible at 1.09 microns, and Pa δ at 1.05 microns.!
(inset). scanning mode. In principle we should obtain an absolute flux calibration
2•10
of less than 3%!
!
0
0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
Spectra of Vega are obtained with spatial scanning in the -1st and,
Wavelength (microns)
depending on position on the array, the -2nd order. !
!
st orders
Calibration
of
-1
!
!
Band 1: 1.15 - 1.32 microns
Band 2: 1.32- 1.4 microns
!
P330E model spectrum
!
Cautions!
-1st! order calibrations. Because use of the -1st grism orders was not anticipated, these were not as
! calibrated as the +1st and +2nd orders. Therefore, we obtained +1st and -1st spectra of the
well
! analog flux standard star P330E to check the flux calibration of the -1st order, and
solar
!
G141 +1st order
observations
of the planetary nebula, VY2-2, to check the wavelength calibration.!
uncorrected spectrum
Band 3: 1.48 - 1.65 microns
1.15 - 1.65 microns, 0th Order
!
!
Upstream/Downstream effects. Scans were made along the columns, either parallel to or opposite
the scan direction. The effect is different scan lengths, and therfore different per pixel on source
exposure times during one image. The WFC3/IR detector is readout using non destructive reads.
DN/sec for V-ALF-LYR
6•108
Flux (electrons/sec)
8
2•108
8
0
-1.8
8
-1.6
-1.4
-1.2
-1.0
Wavelength (microns)
-0.8
-0.6
wfc_coadd 2-Jan-2013 23:14:48.00
8
8
8
8
Flux (electrons/sec)
8
Flux (electrons/sec)
4•108
8
8
8
Flux (electrons/sec)
8
8
8
P330E Spectra
8•10-15
6•10-15
1.034
1.031
1000
0.975
1.036
800
1.002
1.025
800
1.008
1.057
2•10-15
600
1.000
1.000
600
1.000
1.000
0
1.0
1.2
1.4
Wavelength (microns)
1.6
400
1.017
1.029
400
1.041
1.033
1.8
200
0
0
1.4
Wavelength (microns)
1.6
0.6
0.4
400 600
X Pixels
0.2
-0.2
1.0
0.988
1.032
1000
0.461
800
0.997
1.044
800
1.249
600
1.000
1.000
600
1.000
400
1.000
1.000
400
0.264
200
1.158
200
G141 -1st order
uncorrected spectrum
0
0
0.0
1.2
1.4
Wavelength (microns)
1.6
0.998
-1st order
200
800 1000
400 600
X Pixels
1.001
+1st order
0
0
800 1000
200
400 600
X Pixels
800 1000
1.8
Ratio of -1st to +1st order at Postion 1 for P330E
ratio
Figure 1. Top panel: P330E model
spectrum in the IR extended from STIS
spectra. Middle and Bottom panels show
coadded WFC3/IR G141 spectra after flat
fielding, sky subtraction, and dispersion
correction. Flux units are electrons/
second.!
Figure 3. Variation in sensitivity in 3 wavelength
ranges over the array with respect to the spectra in
the middle position. Red points are obtained from
the -1st order spectrum, blue points are from the
+1st order. These are consistent with the results
from Lee et al 2013!
Ratio of -1st to +1st order at Postion 2 for P330E
0.04
0.04
0.03
0.03
0.02
ratio of spectra
0.01
0.00
0.02
0.01
0.00
residual wrt array average
-0.01
-0.01
1.1•104 1.2•104 1.3•104 1.4•104 1.5•104 1.6•104 1.7•104
Wavelength (angstroms)
1.1•104 1.2•104 1.3•104 1.4•104 1.5•104 1.6•104 1.7•104
Wavelength (angstroms)
Ratio of -1st to +1st order at Postion 4 for P330E
0.04
0.04
0.03
0.03
0.02
0.02
ratio
ratio
Ratio of -1st to +1st order at Postion 3 for P330E
0.01
0.00
0.01
0.00
-0.01
-0.01
1.1•104 1.2•104 1.3•104 1.4•104 1.5•104 1.6•104 1.7•104
Wavelength (angstroms)
1.1•104 1.2•104 1.3•104 1.4•104 1.5•104 1.6•104 1.7•104
Wavelength (angstroms)
0.04
0.03
0.03
0.02
0.02
0.01
0.00
0.01
0.00
-0.01
-0.01
1.1•104 1.2•104 1.3•104 1.4•104 1.5•104 1.6•104 1.7•104
Wavelength (angstroms)
Figure
2.
Derived
sensitivity
functions for the two orders. Black
curves are the values at launch.!
Average Ratio over all Positions for P330E
0.04
ratio
ratio
Ratio of -1st to +1st order at Postion 5 for P330E
Left Figure: Measured length of a scan (in pixels) during one read when the scan direction and the
readout direction are the same (parallel), and when the scan and readout direction are opposite each
other.!
!
Right Figure: Ratios of measured parallel and opposite scan lengths to expected length if readout
time were instantaneous. Best fit lines are forced to include the values at scan rate= 0 arcsec/
second. 200
1000
1.8
0.8
electrons/sec
800 1000
Y Pixels
1.2
Y Pixels
0
1.0
!!
! The measured scan
! length is a function of
! scan rate, the
the
! sample time (time
! between reads) and the
! time to readout the
! array. Our spectra were
!obtained using the
!RAPID sample sequence
which has uniform
intervals of 2.93
seconds between reads.
The time to readout one
quadrant is 2.91
seconds. !
400 600
X Pixels
0
0
1.006
+1st order
0.981
-1st order
40
20
Upstream/Downstream Effect!
200
200
0.995
+1st order
ratio
electrons/sec
60
1.003
-1st order
Y Pixels
4•10
1000
-15
Y Pixels
Flux (ergs/s/cm^2/Angstrom)
1•10-14
1.1•104 1.2•104 1.3•104 1.4•104 1.5•104 1.6•104 1.7•104
Wavelength (angstroms)
Figure 4. Ratios of the -1st to +1st spectra at!
each of 5 positions, with residuals (blue curves)
with respect to the average shown in the lower
right panel. On average, the throughput of the
-1st order is 100 times less sensitive.!