Explanatory Notes for the Full-Resolution Flux and Polarization Spectra
Gzipped Tar files containing the full-resolution flux and polarization
spectra for objects observed during the monitoring program can be downloaded
from http://james.as.arizona.edu/~psmith/Fermi/DATA/specdata.html
for data obtained after the beginning of Fermi Cycle 2
(2009 August).
Spectra obtained in Cycle 1 (2008 Oct--2009 May) can be found using
the links to individual objects at
http://james.as.arizona.edu/~psmith/Fermi/DATA/individual.html.
All files are FITS.
All spectra are in the observed
frame of reference.
Flux Spectra:
- The spectra are identified using the following conventions:
- <object>.spec_<UT date>.fits
- <object>.fc.spec_<UT date>.fits
The UT year-month-day of the observation
date is given in the format of yymmdd.
- Objects that have had multiple spectra obtained during a single night are
denoted by the convention: <object>_1..., <object>_2..., etc.,
corresponding to the order of the observations.
- The spectra span the range of 4000-7550 Å with a dispersion
4 Å/pixel.
Depending on the width of the slit used for the observation, the resolution
is typically between 16-24 Å.
- Slit width, exposure duration, airmass and epoch of the observation can
be found in the FITS header. The 'APERTURE' values correspond to the following
slit widths: 1 = 2", 2 = 3", 3 = 4.1", 4 = 5.1", 5 = 7.6", 6 = 12.7".
- Flux density, Fλ, is given in erg cm-2 s-1 Å-1. Wavelength, λ, is given in Å.
- The atmospheric O2 B-band absorption feature at about
6980 Å has not been removed from the spectra.
On particularly humid nights, the H2O atmospheric
absortion feature at about
7200-7300 Å can also be seen.
- At wavelengths longer than about 7000 Å, fringing by the thinned
CCD of SPOL is apparent in most spectra. As with the atmospheric
absorption features, the fringing does not corrupt the polarimetry and
so the spectra provided are extended out in the red to where blue light
from the 2nd-order spectrum begins to contaminate the first order spectrum.
Because properly correcting the fringing of the detector would consume
a substantial
amount of time at the telescope, we have opted not to make this correction
for the flux spectra.
- Files containing ".fc." in the filename denote spectra that have
been scaled to match the V-band magnitude determined from the
differential spectrophotometry on that night.
Specifically, the flux density averaged between
5400-5600 Å in the spectrum
has been scaled to agree with the flux density
determined from the synthetic V-band photometry.
The adopted conversion from flux density
(in mJy = 10-26 erg cm-2 s-1 Hz-1)
to V apparent
magnitude is
given by V = 15 - 2.5 log (Fν/3.64).
- Flux density levels for the spectra without ".fc."
in the filename should NOT be used as true measurements of an
objects's brightness, though the spectral shape is as valid,
having been corrected for
airmass (using the standard KPNO extinction curve within IRAF)
and the instrumental sensitivity.
- No corrections to the spectra have been made for Galactic interstellar
extinction and reddening.
Polarization Spectra:
- The full-resolution Stokes spectra are identified using the
following conventions:
- <object>.q_<UT date>.fits
- <object>.u_<UT date>.fits
The UT year-month-day of the observation
date is given in the format of yymmdd.
- Objects that have had multiple measurements
obtained during a single night are
denoted by the convention: <object>_1..., <object>_2..., etc.,
corresponding to the order of the observations.
- These "q" and "u" spectra are the normalized linear
polarization Stokes parameters as a function wavelength (in Å).
- The Stokes parameters are rotated from the instrumental frame
to the usual astronomical reference frame
such that the polarization position angle (θ) is given by 0.5 tan-1 (u/q),
where θ = 0º when u = 0 and q > 0 (N);
θ = 90º when u = 0 and q < 0 (E);
θ = 45º when q = 0 and u > 0;
θ = 135º when q = 0 and u < 0.
- The percentage linear polarization is given by P =
SQRT(q2 + u2), modulo a correction
needed to account for the non-normal error distribution of P
(see e.g., Wardle & Kronberg 1974).
- The spectra span the range of 4000-7550 Å with a dispersion
4 Å/pixel.
Depending on the width of the slit used for the observation, the resolution
is typically between 16-24 Å.
- Slit width, exposure duration, airmass and epoch of the observation can
be found in the FITS header. The 'APERTURE' values correspond to the following
slit widths: 1 = 2", 2 = 3", 3 = 4.1", 4 = 5.1", 5 = 7.6", 6 = 12.7".
- No corrections to the spectra have been made for Galactic interstellar
polarization.
Please contact Paul Smith (pssmith@arizona.edu) if you have questions concerning these web pages.