BCD processing has two main steps: (1) calculation of the slope of the data ramp (SLOPER) and (2) calibration of the slope image (CALER). MIPS-Ge raw data are comprised of data ramps of 24, 32, or 80 non-destructive reads (for 3, 4, 10 MIPS-sec data collection events (DCEs) respectively).
The namelist file controlling "SLOPER" processing is cdf/{2/3}/MIPS{70/160}_SLOPE_0.nl. The first step in the SLOPER processing of the data is CVTI2R4g which converts the input integer values into real values and makes the initial dmask and bmask files (output "dce.fits"). The SATURATION module sets bits in the dmask for saturated samples. The ELECNL module corrects for electronic nonlinearity (output "enlcorrd.fits"). RESET checks for extra resets in the DCE and sets the dmask appropriately. RADHIT checks for radhits in the ramps and sets the dmask at the location of radhit(s). The SSC pipeline identifies discontinuities using a maximum likelihood technique (Hesselroth et al. 2000, Spaceborne Infrared Remote Sensing VIII, SPIE Conference Proceedings, 4131, 26). SLOPE calculates the linear slopes of segments with at least 4 consecutive good reads as indicated by the dmask (output "currents.fits"). FUSION calculates a noise-weighted average slope from the segments based on the empirical errors estimated from the scatter of the data within the ramp segments (output "current.fits"). The slope image (current.fits) is the final product of the SLOPER part of the pipeline. The last step in sloper is MASKWRITE which copies information from the dmask and pmask to the bmask file for use in the 2nd part of the pipeline.
The 2nd part of processing [CALER] is controlled via the namelist file cdf/{2/3}/MIPS{70/160}_FLUXCAL_0.nl. Calibration and filtering are done on data cubes. Input cubes are made by stacking the SLOPER products with the STACKLAYER module. The calibration of MIPS-Ge data is based on frequent measurements of internal stimulator flashes (stims), which are used to track the response of the detectors as a function of time. The stim flash signal is measured by subtracting the previous ``background'' DCE from the stim frame, which is taken at the same position on the sky. For each AOR, a stim response function (SR[t]) is calculated from interpolating between the stim minus background measurements, which is done via the module INTERP (output "interpstim.fits"). After the determination of the stim response as a function of time (SR[t]), the BCD data are calibrated in SLOPECAL module using the following equation:
Equation 1.1
where U(t) is the uncalibrated slope image, DARK is the dark calibration file, and IC is the illumination correction calibration file which corrects for the combined illumination pattern from the telescope and the stim flash signal. The DARK and IC calibration files are stable and are generated by combining data from several different campaigns to improve the signal-to-noise. The flux conversion factor (FC) converts the instrument units into physical surface brightness units of MJy/sr and is derived from observations of standard calibrators.
The SLOPECAL module of CALER, which applies the calibration, also does optional filtering. The online filtering process is optimized for point sources. Filtering removes data artifacts, but also removes extended source flux. Examples and discussion of MIPS-Ge artifacts are presented in the MIPS Instrument Handbook. The two main artifacts impacting Ge data are the stim flash latents and the variations of the slow response as a function of time. The 160 micron data are affected by these issues to a lesser degree due to the faster time constants of the 160 micron stressed-Germanium detectors. For point sources in low backgrounds (e.g., Frayer et al. 2006, AJ, 131, 250), the stim latent and slow response residuals are additive effects. The Ge filtering removes a running median per pixel as a function of time by subtracting the median value of the surrounding DCEs closest in time (ignoring the current DCE, stim DCEs, and bad data). This "high-pass" time filter does not remove all of the affects of stim latents. To remove the stim latent at 160 micron, one can throw out the first DCE after the stim flash since the stim latents typically decay away within one DCE. At 70 micron stim latents remain for many DCEs and are correlated by column. Since the scan map direction and photometry scan dithers are nearly along the columns of the array, the column artifacts are amplified. Column residuals at 70 micron are removed by subtracting the median of the values along each column for every DCE. The combination of the high-pass time median filter per pixel and the column median filter removes the bulk of the data artifacts at 70 micron.
1.8 GeRT Products
The GeRT makes many intermediate products in addition to the BCD products one can retrieve from the online archive. These intermediate products provide important diagnostic information about the data as well as enabling the testing and optimization of the pipeline software. The products from SLOPER (the first part of the pipeline) are stored under OUTdir/0XXX, where XXX = increasing number corresponding to the files in the raw input list.
1.8.1 SLOPER Products:
bmask.fits bmask file used to identify bad pixels (32x32)
current.fits Final slope image from SLOPER
current_unc.fits Uncertainty of slope image
currents.fits Slopes of line segments within the ramps
currents_unc.fits Uncertainties on the line segment slopes
dce.fits Raw data ramp converted into real
dmask.fits Mask used to identify bad reads in ramps (32x32xN)
enlcorrd.fits Data ramp corrected for the electronic nonlinearity
enlcorrd_unc.fits Uncertainty on enlcorrd.fits
sloper.log Log file for the sloper processing
The intermediate products of the CALER part of the pipeline are stored in OUTdir, while the BCD products are stored in the OUTdir/bcd directory.
1.8.2 CALER Intermediate Products:
SCLK_OBS.fits Vector giving DCE timing information used for stim interpolation (SCLK_OBS keyword)
*.txt Output listing of file names for GETLAYER
*list Input listing of file names for STACKLAYER
bmaskcube.fits Final bmask cube from SLOPECAL module associated with nofiltcube
cal@ Symbolic link to cal files ($SOS_GeRT/cal)
calcube.fits Final filtered data cube from SLOPECAL
calcube_bmask.fits Mask file associated with calcube
calcube_unc.fits Uncertainty file associated with calcube
curcube_unc.fits Uncertainty associated with curcube
darksubt.fits Intermediate file used for making the IC in the calibration pipelines [U(t)/SR(t) - DARK]
darksubt_bmask.fits Mask file associated with darksubt
darksubt_unc.fits Uncertainty file associated with darksubt
fluxcal.nl Copy of MIPS*_FLUXCAL_0.nl namelist file used by CALER part of the pipeline
interp_bmaskcube.fits Bmask file output from INTERP module
interpstim.fits Stim Response cube from INTERP
interpstim_unc.fits Uncertainty file associated with interpstim
nofiltcube.fits Unfiltered BCD cube
nofiltcube_unc.fits Uncertainty file for nofiltcube
slopecal.nl Copy of MIPS*_SLOPE_0.nl namelist file used by SLOPER part of the pipeline
1.8.3 BCD Products (OUTdir/bcd directory)
bcd.fits Unfiltered BCD products made from nofiltcube.fits
bunc.fits Uncertainty file made from nofiltcube_unc.fits
bmask.fits bmask file made from bmaskcube.fits
fbcd.fits Filtered BCD product made from calcube.fits
The bunc and bmask files are applicable for both the BCD and fBCD products. These suffixes match the names of the SSC archived product names. The full name of the GeRT BCD products are given by OUTdir.XXXX.sufix.fits, where OUTdir is the user given output directory, XXXX is an increasing number associated with the DCEs in the input list (e.g., 0001 is associated with the first DCE in the list, 0002 associated with the 2nd DCE...), and sufix = (<=5)-character suffix (bcd, bunc, bmask, fbcd).
Please note if combining data with MOPEX, all input files names must be unique (recommend using different OUTdir names for different AORKeys when working with large data sets).
1.8.4 Calibration Pipeline Products
For the DARK pipeline (EXPTYPE=d2a), the products are (OUTdir directory):
darkfinal.fits Output DARK calibration file corresponding to online calibration file MIPS*_DARK.fits.
darkfinal_mask.fits Mask file for DARK corresponding to online file MIPS*_DARK_C.fits.
darkfinal_unc.fits Uncertainty file for the DARK corresponding to online file MIPS*_DARK_U.fits.
MIPS*_DARK_M.fits Coverage map of DARK file (representing number of "good" DCES used per pixel)
For the IC ("flat") pipeline (EXPTYPE=sfl,pfl,dfl,tfl,ffl), the products are (OUTdir directory):
ilcorr.fits: Output IC calibration file corresponding to online calibration file MIPS*_ILCORR.fits.
ilcorr_cmask.fits: Mask file for the IC file corresponding to online file MIPS*_ILCORR_C.fits.
ilcorr_unc.fits Uncertainty file for the IC file corresponding to online file MIPS*_ILCORR_U.fits.
MIPS*_ILCORR_M.fits Coverage map of IC file (representing number of "good" DCES used per pixel)
