The dark recipe generates one single product:

Master dark frames of the Hawaii array in NonDest read mode.

Purpose. Dark frames are measured to monitor the overall performance and status of the Hawaii array in NonDest read mode.

Recipe. The pipeline input stack consists of three raw frames coherent in the detector setup (meaning the same readmode and discrete integration time = DIT). The recipe generates the master dark frame and calculates several quality control parameters, like dark current and read out noise per detector quadrant.

Note: darks for spectroscopy are processes in the same manner as darks for imaging; the frames differ only in the detector read mode, that is double correlated read mode for imaging and non-destructive for spectroscopy.

The pipeline recipe generates one single product:

Spectroscopic master flat products for the LR-grating
upper left: SZ filter 1.06 micron,
upper right: J filter, 1.25 micron,
lower left: SH filter , 165 micron,
lower right: SK filter, 2.20 micron

Spectroscopic master flats for the median resolution grating (MR-grating)
upper left: SZ filter, 1.06 micron,
upper right: J-filter, 1.25 micron,
lower left: SH-filter , 165 micron,
lower right: SK-filter, 2.20 micron.

Purpose:

The master spectroscopic flat collects instruments signatures to be removed. Among these there are the pixel-to-pixel variations in the spectroscopic read mode, the slit function, interference patterns like fringes and instrument internal reflections in the optical path. For central wavelengths close to the IR bands, telluric absorption lines features become visible in lamp flats. The wavelength position of both ISAAC gratings LR and MR are only reproducible with 2-3 pixel. Therefore night time science spectra and telluric standard star spectra and their imprinted instrumental signatures differ slightly from those recorded in the day-time spectroscopy flats. Science programs that require a high signal-to-noise (>100) should request night-time flats. The science star and the telluric standard star should be reduced using the same spectroscopic flat.

Recipe. The recipe expects one or more pairs of frames, one of the frame being an lamp-off frame the other a lamp-on frame. The day-time spectroscopic flats come as a stack of three pairs, the user-requested attached night time flat calibrations contain one pair of frames. The recipe subtracts pair wise the off-lamp frames from the on-lamp frames and generated the normalized average. In a further reduction step the rims of the flat spectrum are cut at certain threshold values.

Note: ISAAC is operated with four fixed wavelength settings when using the low resolution (LR) grating. For the medium resolution grating the wavelength is a free parameter.

The pipeline recipe generates two products::

Here are two example ARC_CORRECT product frames using the low resolution grating:

Spectroscopic arcs line frame products. The off-lamp frame has been subtracted and the optical distortion has been applied to straighten the lines.
Left: SK-filter , 2.20 micron, Argon lamp spectrum,
Right: SK-filter, 2.20 micron, Xenon lamp spectrum

Spectroscopic arc line product using the medium resolution grating.
Left: SZ-filter 1.06 micron, Xenon+Argon lamp spectrum,
Right: J-filter, 1.25 micron Xenon+Argon lamp spectrum,

Here is an example ARC_COEF pipeline product table:

#
# file           IS_PACF_070915A_SW_LR_SK_2.2_sl1_Ar.fits
# extensions     1
# --------------------------------------------
# XTENSION       1
# Number of columns 4
#
Degree_of_x|Degree_of_y|poly2d_coef|WL_coefficients
          0|          0|   -1.04619|        18437.9
          1|          0|    1.00743|         7.1857
          0|          1|-0.00483497|   -5.99943e-06
          1|          1|-1.0234e-05|   -1.21804e-08
          2|          0|-1.23358e-06|              0
          0|          2|5.99489e-06|              0

The first and the second column are the degree of the polynomial x and y coefficients respectively. The third column contains the coeffients of the the optical distortion (read -1.04619 + 1.00743*x -0.00483497 * y -1.0234e-5 *x*y ...). The fourth column contains the coefficients of the wavelength dispersion relation: wlen=18437.9 + 7.1857*x -5.99943e-6*x*x -1.21804e-8*x*x*x. .

Purpose. Spectroscopic arcs are taken for two reasons:

1. The position of the arc lines are used for wavelength calibration.

2. The curvature of the arc lines provide information about the optical image distortion in dispersion direction (= x-direction). The isaac_spc_ jitter recipe therefore requires an arc product at least once to correct for the optical image distortion and optionally a second time for the wavelength calibration.

Recipe. For reasons to avoid blends in crowded arc line spectra and on the other side to avoid blank spectral regions without any emission line, arc frames are either recorded with an Argon lamp or a Xenon lamp in separate exposures (most settings of the low-resolution grating) or frames where both lamps are on simultaneously (most settings using the medium resolution grating). The pipeline recipe handles all cases.
Although this operational flexibility, there are some setttings, for which the number of found arc lines is not large enough to perform both optical distortion and wavelength dispersion solution. These settings are listed here.

The pipeline recipe generates three product tables:

A combination of 11 stellar raw spectra and the 11 star images subtracted, to demonstrate the Y-shift of the spectrum versus the target position in imaging. The MR grating has been used.

Same as above for the LR-grating.

Here is an example extraction of the STARTRACE_COEF product table:
The third row shows the optical distortion coefficients.

#
# file           IS_PSTC_060117A_SW_LR.fits
# extensions     1
# --------------------------------------------
# XTENSION       1
# Number of columns 3
#
Degree_of_x|Degree_of_y|poly2d_coef
          0|          0|   -10.6467
          1|          0|  0.0176449
          0|          1|    1.00641
          1|          1|-4.77351e-06
          2|          0|-3.37884e-06
          0|          2|-1.12436e-06

Here an example extraction of the STARTRACE_POSI product table.
The first column is the y-pixel on which the star was recorded in imaging.
The second columns is the y-pixel on which the spectrum was recorded.

#
# file IS_PSTP_060117A_SW_LR.fits
# extensions 1
# --------------------------------------------
# XTENSION 1
# Number of columns 2
#
Star_positions|Spec_positions
 897.145| 885.619
 822.162| 811.829
 747.322| 738.117
 672.811| 663.176
 598.648| 588.779
 524.226| 513.911
 449.947| 438.311
 375.279| 363.286
 300.759| 287.3
 226.642| 211.277
 151.578| 134.344

Here an example extraction of the STARTRACE_SHAPE product table.
For each of the 11 calibration spectra, the polynomial coefficients describing the
spectrum curvature are given:

#
# file           IS_PSTS_060117A_SW_MR.fits
# extensions     1
# --------------------------------------------
# XTENSION       1
# Number of columns 11
#
      Spec_1|      Spec_2|      Spec_3|      Spec_4|      Spec_5| ... 
     152.179|     230.054|     306.916|     383.619|     459.856| ...    
   0.0238853|   0.0201866|   0.0174566|   0.0157797|   0.0133045| ...  
-1.40377e-05|-1.02427e-05|-6.92649e-06|-6.18515e-06|-3.82259e-06| ...
 1.64035e-09| 7.66724e-10|-2.87712e-10| 1.88146e-10|-3.69285e-10| ...

Purpose. The startrace template provides three sets of 11 exposures. The first 11 frames are in imaging mode to find the position difference between star image and related spectrum position. The second set consists of 11 raw spectra of a bright star with equidistant offsets along the slit for the LR grating. The third set of eleven spectra are acquired using the MR grating. The grid of stellar spectra is used to derive the optical distortion in spectroscopy in y-direction. The tilt and the curvature of the spectra are achromatic

Recipe. The recipe takes a stack of 3 times 11 raw input frames and produces six product tables, three ones for the LR-grating and three ones for the MR-grating. The STARTRACE_COEF product table is required in further steps of the calibration cascade; this product is delivered only.

The pipeline recipe generates a product frame and a product table:

This image shows the sum of both raw frames.

Telluric standard star final pipeline product. Both spectra have been flat fielded, optical distortion corrected, and co-added. The spectroscopy setup is: SW-arm, LR-grating, SK-filter 2.2 micron, and 1 arcsec slit.

Extracted spectrum of the telluric standard star, as recorded in the SPEC_EXTRACTED product table. The spectroscopic setup is SW-arm, LR-grating, 2.2 micron, SK-filter, 1arcsec slit.

Purpose.

Telluric standard stars contain beside the stellar features of the spectrum many telluric absorption lines, which are imprinted in the spectrum of the science target as well. The telluric standard star is usually taken directly after the science spectrum with the same spectroscopic setup and the same air mass to avoid significant changes of the telluric absorption lines in shape and intensity. When the science target is reduced with the same spectroscopic flat as the telluric standard star, the lamp features cancel out. Telluric standard star products are stand-alone deliverables. No spectral de-composition of the telluric and the stellar features is implemented in the pipeline.

Recipe. The isaac_spc_jitter recipe takes

• a stack of raw telluric standard star frames,
• a static calibration CALPRO_OH_CATALOG containing the OH-line list.
• a spectroscopic flat,
• a startrace product table of type STARTRACE_COEF for the optical distortion in direction perpendicular to dispersion.
• an arc product table of type ARC_COEF for the optical distortion in dispersion direction.

The --wavecal=std option is an important operational change. While ISAAC products (telluric standard stars and science spectra) taken in 2005 and earlier were wavelength calibrated using the day-time arc frames, observations taken in 2006 and later are calibrated using the imprinted sky emission lines. The advantage is, that the low grating position reproducibility (= unavoidable grating offsets between night-time observations and day-time calibrations) does no longer impact the dispersion solution. The drawback is, that standard stars as bright sources with low DIT values show only faint sky lines, hence the dispersion solution might become not quite accurate. The cross-correlation value of the dispersion solution should be higher than 45%. Spectroscopic science products with higher DIT and hence stronger sky emission lines are less affected by low quality dispersion solutions.

Note An Atlas of OH lines is given by Lidman & Cuby, 2000

The pipeline recipe generates a product frame and a product table:

Recipe: The spectra taken in slit less spectroscopy; the raw frames do not show sky lines. They are corrected for optical distortion as in the case of telluric standard stars. The arc product must be used for wavelength calibration as there are no sky lines. The frames are not flat fielded to avoid contamination by features intrinsic to the halogen lamp. The extracted spectrum is compared with library spectra to derived the efficiency and the conversion of the spectroscopic band pass.