Archive Pipeline Products

IRS Pipeline Data Products and Filenaming Convention

Which archive products should you download?

IRS Archive Products

You will most probably want to begin downloading just a few products, and then investigate the details of other files, if necessary. We provide here an overview of the pipeline and archive products to guide you through the archive products and help you decide which products to select for download for your scientific purpose:

Each IRS data collection event (DCE) available from the archive comes with 19 different data products. On top of this there are calibration files, coadded products etc., whose number varies with observation (AOR). You can begin with just a few, and then investigate the details in other files if necessary.

For a first look at the data, begin with the *bcd.fits and *spect.tbl files. These are the pipeline final products: the 2D dispersed images and extracted spectra respectively. Also look at the *wwave.fits file which gives the wavelength associated with each pixel. The *func.fits files provide the uncertainties associated with the bcd.fits images; the 1D spectra tables contain their own uncertainties. The *bmask.fits files contain bit-flags to identify pixel status for the *bcd.fits images.

If your observation included an IRS peakup, that data is also available in *bcd.fits files, although they are mostly processed onboard the spacecraft, not by the science pipeline. See the peakup FAQ for more information.

Naming Convention

Files have one of two basic naming conventions:

SPITZER_S+module_aorkey_expid_dcenum_version_type.suffix
or
SPITZER_S+module_productid_version_type.suffix

The second naming scheme is used for files which do not belong to a single data collection event, e. g. coadded images.

where Module is:
S0, S1, S2, S3 correspond to short-low (SL), short-high (SH), long-low (LL), long-high (LH)
S5, S6 correspond to peakup only blue (PB), peakup only red (PR)
In cycle 2, PU imaging will be introduced and we will have: S8, S9 correspond to peakup imaging blue (PBI), peakup imaging red (PRI) Note that the files called *bcd.fits can be the product of different observations: peakup image, spectrum data or calibration data. The processing steps that generated these files can be found in the HISTORY portion of their header.

FITS Header Keywords

There are many FITS header keywords. They will be described in detail in the IRS Data Handbook. Here we list a few that may be of immediate interest:

1) FOVNAME = The name of the field of view for the observation

For example, a staring mode observation in Short-Low order 1, will take frames in both 'IRS_Short-Lo_1st_Order_1st_Position' and 'IRS_Short-Lo_1st_Order_2nd_Position'

There are several integration time related keywords. The one to look for is:

2) RAMPTIME = Total integration time for the data

For book keeping, it may be useful to search for particular targets by looking at:

3) OBJECT = Target Name

The RA/DEC/position angle of the observation is recorded in several ways:

4) RA_RQST, DEC_RQST, PA_RQST = The requested value for the current frame is given in the keywords

Here the position angle is given for axis 2, counted eastwards from North. Its uncertainty can be found in they keyword UNCRTPA.

If the observation is part of a group (e.g. "cluster mode offsets") then the central target position may be different than the requested position of the current frame:

5) RA_REF, DEC_REF = The original target position

The reconstructed pointing of the telescope is recorded for both the requested field of view and the center of the slit. So in a staring mode observation of a single target, the pointing on the center of the slit will change at the two nod positions, but the two fields of view will each be pointed at the same position (with some small pointing uncertainty):

6) RA_FOV, DEC_FOV, PA_FOV = The field-of-view pointing

7) RA_SLT, DEC_SLT, PA_SLT = The center of the slit pointing
(NB: S11 processed IRS calsfx peak-up data do not have the keywords RA_SLT, DEC_SLT, PA_SLT, but the spectra do.)

8) PTGDIFF = The difference between the reconstructed pointing (for the field of view) and the requested pointing

The data units are given in:

9) BUNIT = 'e-/sec' / Units of image data

Brief Pipeline Summary

The pipeline includes many steps, and the details of the processing will be described in a separate chapter of the IRS Data Handbook. Here we provide only a quick description (see the Spitzer Observer's Manual for a processing schematic).

The pipeline begins with RAW mode data taken in "ramps" consisting of multiple non-destructive reads. The number of reads depends on exposure time. The unprocessed data cube is the *raw.fits file. Initial processing steps performed on the 3D data include: saturation flagging, dark subtraction, and linearity correction. The processed, three-dimensional data cube is the *lnz.fits file. The pipeline rejects then cosmic rays, calculates the slope of the ramp for each pixel and corrects for some detector behaviors, producing the 2D dispersed image *droop.fits. This is the first slope image available. At this point the pipeline splits into two alternate paths: one which includes a correction for stray light (SL) and order-to-order crosstalk (SH, LH), producing the two-dimensional file rsc.fits, and the other doesn't. The final step in the processing of 2D images is flatfielding, producing either the *bcd.fits file, after rsc, or the *f2ap.fits file, without stray light/crosstalk correction. Reduced 2D data for observations within an AOR taken at common pointings in a single FOV are coadded using a weighted average based on source flux (*coa2d.fits). 1D spectra are extracted from the individual bcd.fits and *f2ap.fits files (in *spect.tbl and *spect2.tbl) and from the *coa2d.fits files (*tune.tbl).

Projected Scales

Go back to IRS Data Products page