Contents: IMG | MOS | IFU

Typical observation and reduction problems are listed here.

VIMOS SCIENCE data have been processed by the pipeline with the best available calibration data. Please note that ESO does not assume any responsibility with respect to the usefulness of reduced data. The adopted reduction strategy may not be suitable for the scientific purpose of the observations.

DPR CATG = SCIENCE
DPR TYPE = OBJECT
DPR TECH = IMAGE or IMAGE,PRE (for pre-imaging)

reduced image of NGC5128 in quadrant 3 taken with filter B in stare mode. Vignetted regions at borders are visible

Recipe

The pipeline recipes vmimobsstare and vmimobsjitter do bias subtraction with overscan removal, flat-fielding and removal of cosmic raysl. Images taken in UBVRI filters are also photometrically calibrated. Combination of images taken in jittered mode is pipeline-supported but not applied for data reduction that is used for producing data packages. Photometric calibration consists of writing in the header of the reduced image the zeropoint keyword (in ADU/sec): "HIERARCH ESO PRO MAG ZERO", with its RMS: "HIERARCH ESO PRO MAGZERO RMS". The zeropoints are in general obtained every night, but when the night determination is not available the closest in time zeropoint is used. For the zeropoints determination see the zeropoint page . The pipeline recipe also converts the header keywords "HIERARCH ESO PRO SKY CCD" containing the sky to CCD transformation into the "WCS" (World-Coordinate-System) format readable by the "Mask-preparation Software" and by Skycat. (The converted keywords are called "CO_{i_j}"). The pipeline recipe also computes qc parameters (see below) and writes them in the reduced image header.

Recipe for pre-imaging

The recipe vmimpreimaging is only used for quick-delivery data, and for this purpose standard (not the latest) master calibrations are used. Pre-imaging is taken in stare mode only. The pipeline recipe vmimpreimaging does bias subtraction with overscan removal, flat-fielding and converts the header keywords "HIERARCH ESO PRO SKY CCD" into the "CO_{i_j}" keywords readable by the "Mask-preparation Software". Photometric calibration is not performed and QC parameters are not computed.

Products

* coded as HIERARCH.ESO.PRO.CATG in the fits header
** index of the PIPEFILE name, coded as PIPEFILE in the fits header
*** used in the delivered name

QC checks

As part of the quality control the following parameters are computed by the pipeline imaging recipe and their value are stored in the header of the reduced image. More information on the qc parameters can be found on the QC pages.

Image quality is computed as the mode of FWHM of all the objects in frame classified as stars by SExtractor, when at least 10 stars are found.

Error on image quality is the sigma of FWHM of the objects classified as stars. It is computed iteractively, excluding from the distribution outliers at more than 3 sigma.

Sky background level is estimated by computing the median pixel value in 100 regions of the frame each of 160x180 pixels, the mean of the 10 lowest values, divided by the exposure time is the estimate of the sky background in (ADU/px/sec).

Error on sky background level is the rms of the 10 lower values selected for the background level estimation.

Limiting magnitude in frame is computed as the position of the most populated bin of the star magnitudes histogram. Binsizes from 0.1 up to 1 magnitudes are tried until the most populated bin contains at least 20 stars. If the 20 stars level for the highest bin is never reached, the limiting magnitude is set to 0.

Error on limiting magnitude is the value of the binsize of the histogram used to determine the limiting magnitude.

DPR CATG = SCIENCE,
DPR TYPE = OBJECT,
DPR TECH = MOS

Raw MOS science frame of quadrant 1 obtained with the LR_blue grism. There are 137 science slits and 3 reference slits (two of them are visible at the left side and one at the right). Cleary visible are zero and -1 (minus one) order contaminations

Recipe MOS

The pipeline recipe vmmosobsstare and vmmosobsjitter apply bias subtraction with overscan removal. Flat-field correction is available but not executed during standard reduction for PI data packages. Observations in the red part of the spectrum show strong fringing whereas in the blue reflections are present. Both effects can cause artificial features in science frames during flat-fielding.

The shift in the wavelength direction with respect to skylines is computed and applied to correct for flexures (the skylines used are listed in the header of the input wavelength dispersion solution table under the keywords "HIERARCH ESO PRO SKY WLENi"). The objects in slits are detected, the 2-dimensional slit extraction is done using the distortion models and the inverse dispersion solution read in the wavelength solution table. The median sky level is subtracted and two images containing 2-dimensional object spectra and sky spectra are produced. These images are resampled to a constant wavelength interval as listed in the header keyword CDELT1 (5.27 A/pix for the LR_blue grism), and the wavelength corresponding to the first pixel is listed in the header keyword CRVAL1 (3700 A for LR_blue). Then, in case of jittered data, the 2-dimensional products are stacked. Finally the 1-dimensional optimal extraction is applied (K. Horne1986, Pubs of the Astronomical Society of Pacific,98,609-617) and the result for each slit is stored in one row of the product image.

Products

* coded as HIERARCH.ESO.PRO.CATG in the fits header
** index of the PIPEFILE name, coded as PIPEFILE in the fits header
*** used in the delivered name

Pipeline products obtained using the above MOS science frame. On the left the product MOS_SCIENCE_EXTRACTED is shown. It contains the 2-dimensional reduced spectra of the objects found in each slit, the 1-dimensional extracted spectra are contained in the product MOS_SCIENCE_REDUCED (not shown). On the right is shown the product MOS_SCIENCE_SKY that contains the 2-dimensional reduced sky spectra that have been subtracted to the object spectra. Both images are resampled to the constant wavelength interval 5.27 A/pix and the starting wavelength is 3700 A. Clearly visible again in the sky spectra are second and zero order contaminations. The correspondence between slits in the raw and the reduced frame is that slits from bottom to top in the reduced frames correspond to the slits counted from bottom to top and from left to right in the raw frame.

Some part of the 2-dimensional reduced object spectra MOS_SCIENCE_EXTRACTED is shown here. Eight slits are visible (separated by the black lines) containing object spectra. The slits have been extracted from the raw frame following the curvature model, and they have been resampled to a constant wavelength step (for the LR_blue grism it is 5.27A/pix). On the right is shown the 1-dimensional spectrum of one of the objects obtained by plotting the corresponding row of the product MOS_SCIENCE_REDUCED. The zero order contamination, when saturated, can not be removed, therefore the apparent emission line at 4900A in the extracted spectrum is caused by that, while the emission at 5450A is real.

Some part of the 2-dimensional sky product MOS_SCIENCE_SKY is shown here; corresponding to the object spectra above. The third and the seventh slits (from the bottom to the top) show zero order contamination, while the second slit shows -1 (minus one) order contamination. The bright sky on the blue end of the seventh spectrum is caused by the spectral extraction overlapping the wavelength range of the next multiplexed slit The 2-dimensional extracted sky image also indicates the quality of the wavelength calibration applied to the science frame. This is done by checking the alignment of the skylines in all slits, and by comparing the wavelengths of some of them with their known value. The wavelength range of blue grisms is poor of strong skylines, and they are all in the redder part of the spectrum. Here, the OI lines at 5577.4 and 6299.7,6363.6 and a NaI blend at 5891.6 are visible

The product WINDOW_TABLE refers to the 2-dimensional extracted spectra (MOS_SCIENCE_EXTRACTED and MOS_SCIENCE_SKY) and contains the following columns:

The product OBJECT_TABLE refers to the 1-dimensional extracted spectra MOS_SCIENCE_REDUCED and contains the following columns:

DPR CATG = SCIENCE,
DPR TYPE = OBJECT,
DPR TECH = IFU

Raw IFU science frame of quadrant 2 obtained with the HR_red grism. Redder wavelengths are toward the top. In this image and in all the data obtained with High/Medium Resolution grisms, only the central pseudo-slit is used (i.e. 400 exposed fiber). In Low Resolution, instead, all four pseudo-slits are used (i.e. 1600 exposed fibers).

Recipe IFU

The pipeline recipe vmifuscience is used to reduce a single IFU science exposure. In addition to a raw science exposure the recipe requires a master bias, an inverse dispersion solution file, a fiber relative transmission file, and a file containing information on the tracing of the fibers. The last three files are obtained with the recipe vmifucalib using arc-lamps and flat-field frames which belong to the night-time calibrations associated to the science exposure. The recipe starts with bias subtraction. Then, it refines the input tracing file using the brightest fiber of the science exposure and refines the input inverse dispersion solution file using the position of some identified sky-lines. The spectra are then extracted, wavelength calibrated and resampled to a constant wavelength step (written in the keyword CDELT1). The extracted spectra are also corrected for the relative differences in fiber transmission (using the input relative transmission file) and finally stored in the output image in the usual order: sucessively from pseudo-slit1 to pseudo-slit4 counting fibers from left to right. Flat-fielding and fringing correction are not executed.

The extracted fiber spectra image is then used together with the appropriate IFU table to reconstruct the image of the field of view. The extracted spectra are integrated in a predefined wavelength range, chosen where the spectra are brighter and excluding zero order contaminations, and the value obtained are stored in the corresponding positions on the IFU head.

Products

* coded as HIERARCH.ESO.PRO.CATG in the fits header
** index of the PIPEFILE name, coded as PIPEFILE in the fits header
*** used in the delivered name

A product image IFU_SCIENCE_REDUCED is shown below, obtained with the raw image shown above. The extracted fibre spectra are 400 and the spectrum of the fiber 352 is plotted on the top. The spectra are resampled to the constant wavelength interval listed in the header keyword CDELT1. CDELT1 is 0.58 A/pix in this image taken with the HR_red grism. The spectral signal is given in ADU per wavelength interval, then, to have the signal per Angstrom one should divide the spectrum by CDELT1. Clearly visible are sky emission and absorption lines. The position of some sky emission lines are also used during the QC checks to control the wavelength calibration.

Product image (PRO CATG: IFU_FOV) of the reconstructed field of view of quadrant 2. The reconstructed image of each single quadrant has the dimension of the 4-quadrant field of view (here 40x40) and is filled with zeros in the regions corresponding to the other quadrants (black regions in figure). Here, in High Resolution, only the 400 fibers of the central pseudo-slit are used that correspond to a 20x20 region of the IFU head. In low resolution all 4 pseudo-slits with 1600 fibers are used corresponding to a region 40x40 of the IFUhead. The mapping between the spectra position in the extracted spectra image and the corresponding points of the reconstructed field-of view image is given in the IFU tables.

Reconstructed image of the 4 quadrants field of view. The 4 quadrant image is the product of the vmifucombine recipe, that simply sums the four single quadrant images. The black regions correspond to dead fibres. In high/medium resolution, the 4-quadrants reconstructed image has size 40x40. In low resolution, the size of the 4-quadrant image is 80x80.