GIRAFFE calibration data are mostly taken as daytime calibrations. Flux standard stars and nasmyth screen flats are taken, as part of the calibration plan, for Argus data.
Find information about the GIRAFFE data format, spectral format and the detector here.
The following calibration data are taken as part of the daily GIRAFFE calibration plan. The data types are identified by the DPR TYPE keyword in the FITS headers.
detector status, bias level
dark current (for maintenance); CCD glow (old CCD only)
FLAT (taken with the calibration lamp on the robot arm; this is the regular flat also called 'robotic' flat)
MOS (for Medusa) or IFU (for IFU and Argus)
signal localization and extraction; fix-pattern noise, fringing
FLAT (taken with the nasmyth screen illuminated; this is the 'attached' flat taken during the night, useful for Argus mode because of better illumination)
The ARC-LAMP and FLAT frames split into five fibre system (MEDUSA1/2; IFU1/2;
Argus). They also split into two different gratings (HR and LR) with a set of
standard wavelengths (which are listed here).
BIAS frames are taken in unbinned (1x1) mode, normal, high or low gain.
The flux STD data are regularly taken for the Argus and IFU settings.
In addition, DARK frames are taken about monthly, to monitor the dark current
A series of about 40 IMAGE FLATs is also taken roughly once per month, to
monitor the status of the detector. They come in pairs with exposure times
increasing from 0 to about 280 sec (saturation). Both types of data are not
required for science reduction.
After production and certification, the calibration products are archived.
They are named with
a scheme including type, creation date and relevant setting
parameters. The master calibrations are available for archive users through
the calSelector service.
The processing of GIRAFFE calibration frames requires a cascaded scheme where
the mutual dependencies of products and raw frames are respected. The proper
sequence of all these production steps is called the calibration cascade. It
is described here.
Within that cascade, all products are checked with respect to their quality
and their relationship in time. This means e.g. that a dispersion solution is
not only generated using an arbitrary quality-checked localization solution,
but that specific solution is chosen which is closest in time.
All products created by QC Garching follow the calibration cascade. On the
hand, Paranal-processed science data use
standard solutions created by QC Garching and possibly
several months old.