QC.DID Data dictionary for VIRCAM QC.|Name/version of ESO DID to which QC keywords comply.
QC.DARKCURRENT average dark current on frame [adu/sec].|measured using the median of the pixel values, can later be compared similar darks for trends
QC.DARKRMS RMS noise of combined dark frame [adu].|RMS is defined here as the Gaussian equivalent MAD ie. 1.48*median-of-absolute-deviation from median The RMS can later be compared with library values for darks of the same integration and exposure times.
QC.RESETDIFF_RMS RMS new-library reset frame|measure the RMS of the difference of a new mean reset frame and a library reset frame.
QC.DARKDIFF_RMS RMS new-library dark frame|measure the RMS of the difference of a new mean dark frame and a library dark frame.
QC.PARTICLE_RATE cosmic ray/spurion rate [count/sec/detector].|average no. of pixels rejected during combination of dark frames, used to give an estimate of the rate of cosmic ray hits for each detector. This can later be compared with previous estimates and monitored.
QC.RESETRMS RMS noise in combined reset frame.|variation is defined here as the Gaussian equivalent MAD ie. 1.48*median-of-absolute-deviation from unity after normalising by median level ie. measuring the RMS reset level variation. The RMS can later be compared with library values for troubleshooting problems.
QC.READNOISE readnoise [electron].|measured from the noise properties of the difference in two consecutive dark frames, using a MAD estimator as above for robustness against spurions. The noise properties of each detector should remain stable so long as the electronics/micro-code have not been modified.
QC.FLATVAR RMS variation of flatfield pixel sensitivity per detector [percentage].|RMS is defined here as the Gaussian equivalent MAD ie. 1.48*median-of-absolute-deviation from unity after normalising by median level ie. measuring the RMS sensitivity variation. The RMS can later be compared with library values for troubleshooting problems.
QC.GAIN gain [e/ADU].|determined from pairs of darks and flatfields of the same exposure/integration time and illumination by comparing the measured noise properties with the expected photon noise contribution. The gain of each detector should remain stable so long as the electronics/micro-code have not been modified.
QC.BAD_PIXEL_STAT fraction of bad pixels per detector [scalar].|determined from the statistics of the pixel distribution from the ratio of two flatfield sequences of significantly different average count levels. The fraction of bad pixels per detector (either hot or cold) should not change significantly with time.
QC.GAIN_CORRECTION ratio of detector median flatfield counts to global median [scalar].|the ratio of median counts in a mean flat exposure for a given detector relative to the ensemble defines the internal gain correction for the detector These internal relative detector gain corrections should be stable with time.
QC.LINEARITY the percentage average non-linearity [percentage].|derived from measured non-linearity curves for each detector interpolated to 20k counts (ADUs) level. Although all infrared systems are non-linear to some degree, the shape and scale of the linearity curve for each detector should remain constant. A single measure at 20k counts can be used to monitor this although the full linearity curves will need to be examined quarterly [TBC] to look for more subtle changes.
QC.LINFITQUAL the RMS fractional error in linearity fit [scalar]|Derived by applying the linearity coefficients to the image data that were used to measure them. This is the RMS of the residuals of the linearised data normalised by the expected linear value
QC.SATURATION saturation level of bright stars [ADU].|determined from maximum peak flux of detected stars from exposures in a standard bright star field. The saturation level*gain is a check on the full-well characteristics of each detector.
QC.PERSIST_DECAY mean exponential time decay constant [s].|the decay rate of the persistence of bright images on subsequent exposures will be modelled using an exponential decay function with time constant tau. Requires an exposure on a bright star field followed a series of darks.
QC.PERSIST_ZERO fractional persistence at zero time (extrapolated) [scalar].|determined from the persistence decay behaviour from exponential model fitting. Requires an exposure on a bright star field followed a series of darks (as above)
QC.CROSS_TALK average values for cross-talk component matrix [scalar].|determined from presence of +ve or -ve ghost images on other channels/detectors using exposures in bright star fields. Potentially a fully populated 256x256 matrix but likely to be sparsely populated with a small number of non-zero values of band-diagonal form. This QC summary parameter is the average value of the modulus of the off-diagonal terms. Values for the cross-talk matrix should be very stable with time, hardware modifications notwithstanding.
QC.WCS_DCRVAL1 actual WCS zero point X - raw header value [deg].|measure of difference between dead-reckoning pointing and true position of the detector on sky. Derived from current polynomial distortion model and 6-constant detector model offset.
QC.WCS_DCRVAL2 actual WCS zero point Y - raw header value [deg].|measure of difference between dead-reckoning pointing and true position of the detector on sky. Derived from current polynomial distortion model and 6-constant detector model offset.
QC.WCS_DTHETA actual WCS rotation PA - raw PA header value [deg].|measure of difference between dead-reckoning PA and true position angle of the detector. Derived from current polynomial distortion model and 6-constant detector model effective rotation term.
QC.WCS_SCALE measured WCS plate scale per detector [deg/pixel].|measure of the average on-sky pixel scale of detector after correcting using current polynomial distortion model
QC.WCS_SHEAR power of cross-terms in WCS solution [deg].|measure of WCS shear after normalising by plate scale and rotation, expressed as an equivalent distortion angle. Gives a simple measure of distortion problems in WCS solution.
QC.WCS_RMS robust RMS of WCS solution for each detector [arcsec].|robust average of residuals from WCS solution for each detector. Measure of integrity of WCS solution.
QC.MEAN_SKY mean sky level [ADU].|computed using a clipped median for each detector Sky levels (perhaps not at Ks) should vary smoothly over the night. Strange changes in values may indicate a hardware fault.
QC.SKY_NOISE RMS sky noise [ADU].|computed using a MAD estimator with respect to median sky after removing large scale gradients. The sky noise should be a combination of readout-noise, photon-noise and detector quirks. Monitoring the ratio of expected noise to measured provides a system diagnostic at the detector level.
QC.SKY_RESET_ANOMALY systematic variation in sky across detector [ADU].|robust average variation in background level for each detector, computed by measuring the large scale variation from a filtered 64x64 pixel background grid, where each background pixel is a clipped median estimate of the local sky level. Effectively generates an 32x32 sky level map and computes the MAD [TBC] of these values with respect to the global detector median. Monitoring the non-flatness of this gives a measure of reset-anomaly problems.
QC.NOISE_OBJ number of classified noise objects per frame [number].|measured using an object cataloguer combined with a morphological classifier. The number of objects classified as noise from frame-to-frame should be reasonably constant; excessive numbers indicate a problem.
QC.IMAGE_SIZE mean stellar image FWHM [arcsec].|measured from the average FHWM of stellar-classified images of suitable signal:to:noise. The seeing will obviously vary over the night with time, wavelength (filter) and as airmass^0.6. This variation should be predictable given local site seeing measures. A comparison with the expected value can be used as an indication of poor guiding, poor focus or instrument malfunction.
QC.APERTURE_CORR 2 arcsec [mag] diam aperture flux correction.|the aperture flux correction for stellar images due to flux falling outside the aperture. Determined using a curve-of-growth of a series of fixed-size apertures. Alternative simple measure of image profile properties, particularly the presence of extended PSF wings, as such monitors optical properties of system; also required for limiting magnitude computations.
QC.ELLIPTICITY mean stellar ellipticity [scalar].|the detected image intensity-weighted second moments will be used to compute the average ellipticity of suitable signal:to:noise stellar images. Shot-noise causes even perfectly circular stellar images to have non-zero ellipticity but more significant values are indicative of one of: optical, tracking and autoguiding, or detector hardware problems.
QC.ZPT_2MASS 1st-pass photometric zeropoint [mag].|the magnitude of a star that gives 1 detected ADU/s (or e-/s) for each detector, derived using 2MASS comparison stars for every science observation. This is a first pass zero-point to monitor gross changes in throughput. Extinction will vary over a night, but detector to detector variations are an indication of a fault.
QC.ZPT_STDS photometric zeropoint [mag].|the magnitude of a star that gives 1 detected ADU/s (or e-/s) for each detector, derived from observations of VISTA standard star fields. Combined with the trend in long-term system zero-point properties, the ensemble "average" zero-point directly monitors extinction variations (faults/mods in the system notwithstanding) The photometric zeropoints will undoutbedly vary (slowly) over time as a result of the cleaning of optical surfaces etc.
QC.LIMITING_MAG limiting mag ie. depth of exposure [mag].|estimate of 5-sigma limiting mag for stellar-like objects for each science observation, derived from QCs ZPT_2MASS, SKY_NOISE, APERTURE_CORR. Can later be compared with a target value to see if main survey requirements (ie. usually depth) are met.
QC.FRINGE_RATIO [scalar] Ratio of sky noise before to after fringe fit|A robust estimate of the background noise is done before the first fringe fitting pass. Once the last fringe fit is done a final background noise estimate is done. This parameter is the ratio of the value before fringe fitting to the final value after defringing.