Using the CASSIS spectra

Default spectra

By default, users can download the nod-combined spectrum. The following parameters apply to the default spectra:

  • wavelength grid interpolated on the reference "wavesamp" table,
  • calibrated,
  • defringed.
  • trimmed (LR6 version). See below for more details

As of version LR6, the default spectra are trimmed, i.e., there is no overlap between the spectral orders and the downloaded spectra is the same as the one plotted in the main result page. If you wish to conserve these overlaps, please use the corresponding option in the alternate download page.

How was the background subtracted in my spectra?
The pipeline chooses which background subtraction works best (between by nod, by order, or nothing) for each module/order, depending on the presence of contaminating sources and the detection quality in each case (check the main documentation). The default downloadable spectrum is a combination of the best choices (i.e., the background method might be different for different modules). You can still display the other spectra by clicking on "Extraction & Background" in the main result page.

About the extraction method
While optimal extraction provides the best signal-to-noise ratio possible, the flux calibration is reliable only for point sources. In version LR5 and up, the best extraction method is chosen automatically, and the tapered column extraction method is used instead optimal extraction if the source is partially-extended. Messages will accompany the spectrum displayed to explain the choice.

About the wavelength grid
The default method is to extract the nod spectra on the wavesamp grid, and then combine the 2 nods on this grid. This gives a spectrum with a wavelength grid compatible with the other extraction methods in SMART. Other methods for the wavelength grid calculation are available (see below).

Optional spectra

Other versions of the spectra are available, corresponding to:

  • a specific background subtraction method,
  • the observed wavelength grid instead of the reference "wavesamp" grid,
  • fringes not removed.
  • spectral order overlaps kept.
Please contact the CASSIS team for more details.

What are the available wavelength grids?
We refer the users to the main pipeline documentation or the paper for more information on how the wavelength grid is handled by the pipeline. The available grids are:

Nod spectra to be combined Final (merged) spectrum File naming convention
wavesamp (default) wavesamp (default) spcfw_*
observed (exact) interleaved (full sampling) spcf_*
observed (exact) wavesamp spcfww_*
If signal-to-noise is the most important parameter, the spcfw_ or the spcfww_spectra should be considered. If spectral sampling is sought, the spcf_ spectra should be used.

File naming convention

The spectral files have the following format:

sp{calibration switch}{defringing switch}{wavelength grid parameter}_AORKEY_{type}{trimmed option}.fits

The files begin with sp for 'spectrum', following by letters corresponding to various switches or parameters in the pipeline:

  • calibration switch: c,
  • defringing switch: f,
  • wavelength grid parameter: w: interpolated on the reference "wavesamp" table, ww: nod spectra interleaved and then interpolated on the reference "wavesamp" table.
  • The type of the spectrum can be:
    • the FOVID and order (26.1, 33.3, etc... see list of FOVIDs),
    • the module and order, if nods are combined (SL1, LL3, etc...),
    • the module, if nods are combined and orders are merged (SL, LL).
  • trimmed option: either t (spectral orders trimmed) or u (not trimmed). Available as of version LR6.

Example of a spectrum file: 'spcfw_1234567_SL1.fits' is the spectrum of AORKEY 1234567 for the module SL, 1st order. The flux was calibrated, fringes were removed, and the wavelength grid corresponds to the reference "wavesamp" table.

How to read the spectral files

Spectra are provided in YAAAR FITS binary format and as VO tables. The FITS files can be read with readfits in IDL and can also be imported in SMART.

In Python, some examples:

  • VO Tables can be read with astropy.table:
from astropy.table import Table 
t = Table.read('cassis_vo_spcfw_12622848/cassis_vo_spcfw_12622848_LL1.fits')
  • IPAC tables can also be read with astropy.table:
from astropy.table import Table 
t=Table.read('cassis_tbl_opt_16205568.tbl', format='ascii')
  • YAAAR FITS files can be read with astropy.io.fits:
from astropy.io import fits
import numpy as np
hdu = fits.open('cassis_yaaar_spcfw_12622848t.fits') 
ncols = len(hdu[0].data[0,:])
wdata = np.array(hdu[0].data[:,0])
fdata = np.array(hdu[0].data[:,1])  
  • Unfortunately, YAAAR FITS files are not recognized by Table.read, but to read YAAAR FITS files as Tables, one can do:
hdu = fits.open('cassis_yaaar_spcfw_12622848t.fits') #default is readonly
names=tuple([hdu[0].header['COL*DEF'][x] for x,h in enumerate(hdu[0].header['COL*DEF'])])
names+=('flag4',) #append flag4, bug in the header keywords for low-res, flag number 4 COLDEF is missing
t=fits.BinTableHDU(data=Table(hdu[0].data, names=names), header=hdu[0].header)

The columns are:

version LR5 (and up) spectra

  • 1: wavelength (um)
  • 2: flux density (Jy)
  • 3: total error on flux (Jy). (rms+systematic).
  • 4: rms error on flux (Jy). Statistical error.
  • 5: systematic error on flux (Jy). Related to differences observed between the spectra of the 2 nods.
  • 6: systematic calibration error on flux (Jy). Related to the flux calibration.
  • 7: extended emission flux density (Jy). Filled by AdOpt?.
  • 8: error on extended emission flux density (Jy). Filled by AdOpt?.
  • 9: module number (0: SL1, 1: SL2, 2: LL1, 3: LL2)
  • 10: spectral order (1, 2, or 3)
  • 11: nod (1 or 2)
  • 12: status
  • 13: flag1
  • 14: flag2
  • 15: flag3
  • 16: flag4

version LR4 spectra

  • 1: wavelength (um)
  • 2: flux density (Jy)
  • 3: rms error on flux (Jy). Statistical error.
  • 4: systematic error on flux (Jy). Related to differences observed between the spectra of the 2 nods.
  • 5: systematic calibration error on flux (Jy). Related to the flux calibration. Set at 0 in CASSIS version 4.
  • 6: extended emission flux density (Jy). Filled by AdOpt?.
  • 7: error on extended emission flux density (Jy). Filled by AdOpt?.
  • 8: module number (0: SL1, 1: SL2, 2: LL1, 3: LL2)
  • 9: spectral order (1, 2, or 3)
  • 10: nod (1 or 2)
  • 11: status
  • 12: flag1
  • 13: flag2
  • 14: flag3
  • 15: flag4

For more information on the errors and the error propagation, check this page

Header keywords

There are numerous keywords propagated during the extraction pipeline.

keyword type values comment
CASVE integer   CASSIS version
CASCL string 'cleaned ' cleaning performed
--- CASCL1 integer   maskval used for cleaning
CASCI string 'image combination' image combination performed
--- CASCI1 string 'single DCE' combination type
---   '2 DCEs, one is NaN? dominated' combination type
---   '2 DCEs, the 2 being dominated by NaNs?' combination type
---   '2 DCEs' combination type
---   'more than 2 DCEs' combination type
---   'more than 2 DCEs, all dominated by NaNs?' combination type
CASRB string 'background removed' background removed
--- CASRB1 string 'nod' background removal method
---   'order' background removal method
---   'insitu' background removal method
CASEX string 'extracted' extraction performed
--- CASEX1 string 'nod' background method, duplicates CASRB1
---   'order' background method, duplicates CASRB1
---   'insitu' background method, duplicates CASRB1
--- CASEX2 float   offset in the dispersion direction, ratio of ptgdiffy/0.14*slit_size
--- CASEX3 float   position of other source in the slit
--- CASEX4 integer   background polynomial order for source extraction
CASCF string 'changed fields' changed fields of spectral fits files
CASCA string 'calibrated' calibration performed
CASCA1 string 'RSRF: ' RSRF used
CASCA (tapered only) string 'point-like source' calibration performed
CASCA (tapered only) string 'partially extended source' calibration performed
CASCA (tapered only) string 'extended source' calibration performed
CASDF string 'defringed' defringing performed
--- CASDF1 integer   S/N threshold used for performing defringing
--- CASDF2 integer   number of cycles for IRSFRINGE
--- CASDF3 integer   number of fringes for IRSFRINGE
CASRG string 'wavsamp' wavelength regridded to wavsamp table
CASCSE string 'spectra of DCEs combined' DCE spectra combination for unstable pointings
    'spectra of ExpIDs? combined' ExpID? combination for cluster observations
--- CASCSE1 integer   number of spectra combined
CASCSN integer/string FOVID single nod used
    'interleaved' nod combination method
    'combined on wavsamp' nod combination method

In addition, there are several relevant keywords from the AdOpt? algorithm:

keyword type values comment
S_OPPOS float   source's position (intersect at row 0)
S_OPPOSR float   source's position (relative to nod)
S_WAVE integer 0, 1 wavelength scaling, 0 if regridded, 1 if not (see CASRG)
RA_SM float   RA of extracted source (in degrees)
DEC_SM float   DEC of extracted source (in degrees)

Other products (for specific use, provided per request)

Other products are available by request only. They include the intermediate image products:

  • cleaned images,
  • combined (+cleaned) images,
  • background-subtracted (+combined, +cleaned) images.

Image products can be read in IDL with the readfits procedure. They are in FITS binary format. They can also be imported and read in SMART (see documentation) for analysis.

Local use of the atlas

For projects aiming at analyzing in a systematic way the atlas, a local access can be arranged. There is not yet a standard way to browse through the data, so we encourage users to contact us directly.

Local access to the CASSIS database

 
This site is powered by FoswikiCopyright © 2002-2018 by the IRS Team.
Send feedback. We use Foswiki.