Quality Control and
Data Processing

This page summarizes known problems with historical data of FORS2.

2011-06-01: PMOS lamp flat spectra show for the 1200B grism and in particular for the 600B grism a feature (a ghost, an optical reflection image of the slit) on the blue side of the PMOS spectra. The pipeline does not mask the reflection feature.

	extraction of a G1200B grism PMOS lamp flat spectrum
	extraction of a G600B grism PMOS lamp flat spectrum

2011-06-01: We recently noticed that for unbinned data (1x1) the world coordinate system (WCS) for the lower CCD is wrong. If you need a WCS for the lower chip please use fsmosaic to combine the upper and the lower chip into one frame. This combined frame will have the (correct) WCS of the upper chip.

• At a given retarder plate angle the artefacts move in opposite X/Y direction of 'normal' targets.
• At a given position the artefacts move with increasing retarder plate angle to positive Y, while 'normal' targets remain fixed.
• Additional artefacts appear at retarder plate angle 45.0/67.5.
• The image quality of the artefacts scales with general image quality

The most probable cause for these artefacts are reflections at the retarder plates.

From early 2010 to November 20, 2010 a ring-like feature appears in FORS2 twilight flats, which moves with the rotator angle. In November the LADC was cleaned, which removed the feature. Below an example for the R-band is shown. Similar effects have been observed before (Moehler et al. 2010, PASP 122, 93 and overview). This limits the achievable photometric accuracy to about 5%.

2010-04-16: We recently noticed that at some time during P84 the zero point of the angle of the retarder plate changed. We are still investigating when this happened and how it will affect the results. Until then please do not use the polarimetric pipeline products for scientific analysis as they may provide wrong results

When a sequence of calibration data for different masks is taken, data with a low number of slitlets taken after data with a high number of slitlets would inherit the additional slitlet keywords from the previous observation. This problem has been fixed on November 13, 2006.

Example: Calibrations for a mask with 20 slitlets are taken after calibrations for a mask with 47 slitlets. In that case the calibrations for the mask with 20 slitlets would inherit the following keywords for slitlets 21 to 47 (33 being the number of the slitlet):

• HIERARCH ESO INS TARG1## NAME
• HIERARCH ESO INS MOS1## RA
• HIERARCH ESO INS MOS1## DEC
• HIERARCH ESO INS MOS1## WID
• HIERARCH ESO INS MOS1## LEN
• HIERARCH ESO INS MOS1## POSANG
• HIERARCH ESO INS MOS1## WIDTH
• HIERARCH ESO INS MOS1## SHAPE
• HIERARCH ESO INS MOS1## XPOS
• HIERARCH ESO INS MOS1## YPOS

top Color Equations for FORS2 filters

Q: I have photometric stars taken during the night in R_special filter, but I only have the R magnitude for them from Landolt standard catalogs. So, I need to know if you have the color equation to do the conversion between R magnitudes and R_special magnitudes.

A: The colour terms for the R_special filter can be found in the PHOT_TABLE files (fors2_1453_phot.fits, fors2_1456_phot.fits), that are delivered with the FORS2 pipeline. The term is -0.0095598 and it should be applied with the colour V-R. However, it is a surprise that your data have Landolt standard star fields, because by default the Stetson fields are used for FORS2. Please verify that you need to use the Landolt catalogue. If so, a different SExtractor configuration file for the pipeline is needed, because the magnitudes in the Landolt catalogues are determined differently from those in the Stetson catalogue. The parameters to be changed are

• PHOT_APERTURES=56
• MASK_TYPE=NONE
• BACK_PHOTOTYPE=GLOBAL
• BACK_SIZE=64
• DETECT_THRESH=3