[ ESO ]
ISAAC Quality Control:
Parameters and Trending

CAL | HC | refs | QC
Trending & QC1
trending & QC1
Data Packages
Data Management
QC links:

ISAAC calibration data are pipeline-processed to create calibration products. The quality of the products is measured by QC1 parameters. Most QC1 parameters are obtained by pipeline procedures; some are measured by post-pipeline recipes. All measured QC1 parameters are stored in a database. Their evolution with time is evaluated by the trending process. ISAAC QC1 parameters are routinely measured for all calibration and science products measured in ISAAC pipeline supported modes.

The plots shown here are TRENDING PLOTS . They show in most cases the evolution of the last three months up to day X, for which data have been processed by the pipeline in Garching. Day X is usually 6-12 days behind the observing night due to hard disc delivery between South America and Europe. See here for a simple data flow diagram. QC1 parameters of data still on their way to Europe are monitored via HEALTH CHECK PLOTS. They are based on ftp delivered plain text files and are assesed within minutes. Historic plots can be found on the health check page.

The product naming scheme uses some shortcuts:

As ISAAC is a multi-mode instrument with two detectors, the number of QC1 parameters and the corresponding number of trending plots is rather large. Therefore we provide one menu in two different versions: one per raw type and one per item. Both menus point ot the QC tutorial pages.

Raw Types Menu






Instrument Properties Menu

detector filters/cameras
gratings optical distortion in spectroscopy lamps operations throughput, all components
dark current flatness dispersion and grating position X-direction distortion spectroscopic flat lamp efficiency tw/sky flat flux range photometric zeropoints
RON bad pixel map   Y-direction distortion spectroscopic arc lamp efficiency slit positions, camera image quality
Fixed Pattern Noise           sky brightness
reset anomaly           response function
oddeven effect            



All QC1 parameters extracted from calibration and science products (and in some cases raw frames) are archived in the central public QC1 database.The database is organized as follows: There is for each VLT/I instrument a set of tables. In most cases each table contains the QC1 parameters related to a prodcut type or a raw type, e.g. DARK, or FLAT. Each of the tables contain three sections: a general section with time keys like MJD-OBS and file name keys, a QC section containg the QC1 value for each product type, and the instrument section with the parameters on which the QC1 values depend. E.g. a QC1 value called 'photometric zeropint' depends on the ins parameter 'filter'. For each table there are two interfaces. A plotting interface and an interface to download the QC1 parameters.

The many QC1 parameters listed in the tables are explained in detail on the QC PRODUCT PAGES.

The QC1 database project started for ISAAC in 2003. At the end of 2003 most of the calibration product tables exist and are filled with values typically lasting back to 2002. Further tables have been generated for QC1 paramaters derived from science products. Filling up the QC1 database with further historic data will be a subject of lower priority.


All ISAAC calibration products are quality-checked. While details of this process vary from case to case, some features are generally valid:

  • Newly derived QC1 parameters are displayed in a and outliers are checked for possible assignable cause.
  • Raw frames and master frames are displayed and visually checked for anomalies.Two kind of difference images are generated, a) the difference with the closest in time taken product, b) the difference with a reference master taken some longer time ago. The visual check is to a) check why a QC1 parameter became an outlier or b) check for variations which cannot be paremeterized in a single numerical value.
  • A calibration product is certified when experience shows that it measures the instrumental status in the proper way. This may mean that it complies with the expected result within certain limits. But it may also be true that it represents a valid outlier. E.g., after a realignment of a grating its X and Y shift against a reference frame will differ from the previous values, but this is a valid deviation.