Quality Control and Data Processing

FITS key QC1 database: table, name definition class* HC_plot** more docu

QC.FLUX1.VAL xshooter_telluric..counts1 average counts in first region (ADU) CAL [docuSys coming]

QC.FLUX2.VAL xshooter_telluric..counts2 average counts in second region (ADU) CAL [docuSys coming]

QC.FLUX3.VAL xshooter_telluric..counts3 average counts in third region (ADU) CAL [docuSys coming]

QC.FLUX1.SN xshooter_telluric..sn1 average signal-to-noise ratio in first region CAL [docuSys coming]

QC.FLUX2.SN xshooter_telluric..sn2 average signal-to-noise ratio in second region CAL [docuSys coming]

QC.FLUX3.SN xshooter_telluric..sn3 average signal-to-noise ratio in third region CAL [docuSys coming]

*Class: KPI - instrument performance; HC - instrument health; CAL - calibration quality; ENG - engineering parameter
**There might be more than one.

Scoring&thresholds Signal-to-noise ratio

The investigated wavelength regions have been chosen in that way that the contamination of the signal to noise ratio by telluric absorption features is minimized in the case of telluric standard stars. The regions are given (in nm):

arm 1 2 3

UVB 450-470 510-530

VIS 627-680 745-756 992-999

NIR 1518-1548 2214-2243

Signal-to-noise ratio and the counts of standard star observations is a calibration product quality. It is therefore not scored and not trended. Alone the SN-ratio versus counts relation is plotted.

Algorithm Signal-to-noise ratio

In 2012 and 2013 several new pipeline versions with better handling of the noise introduced a better SNR-to_counts relation.

The signal-to_noise is derived from spectral regions that are not biased by telluric absorption features.

Saturation

Signal-to-noise ratio | Saturation | Efficiency (for flux standard stars only) | Response function (for flux standard stars only)

FITS key QC1 database: table, name definition class* HC_plot** more docu

QC.FLUX1.VAL xshooter_telluric..npix_high (maximum) number of pixels above 60000 ADU (UVB/VIS) in raw frame(s) CAL [docuSys coming]

QC.FLUX2.VAL xshooter_telluric..npix_sat number of saturated pixels in raw frame(s) CAL [docuSys coming]

QC.FLUX3.VAL xshooter_telluric..fpix_sat fraction of saturated pixels in raw frame(s) CAL [docuSys coming]

*Class: KPI - instrument performance; HC - instrument health; CAL - calibration quality; ENG - engineering parameter
**There might be more than one.

Scoring&thresholds Saturation

npix_high gives the maximum number of highly exposed pixels (>60000 ADU for UVB/VIS) in the raw data. These pixels are not saturated but very likely significantly non-linear in their response.
npix_sat gives the number of saturated pixels (65535 ADU for UVB/VIS, 0 for NIR) in the raw data. The correct implementation for NIR data has been done only from version 0.9.8 on, but unfortunately also unsaturated pixels may have a flux of 0 or less in NIR raw data.
fpix_sat gives the fraction of saturated pixels (65535 ADU for UVB/VIS, 0 for NIR) in the raw data. The correct implementation for NIR data has been done only from version 0.9.8 on, but unfortunately also unsaturated pixels may have a flux of 0 or less in NIR raw data.

The detector response non-linearity at saturation level (65535 ADU) has been retrieved from detector monitoring data from 2103-06-13 within a small analysis window of 100x100 pixels ( [x=1500-1600 : y=1700-1800]) with homogeneous flat lamp illumination to be:

arm 400k 1x1 100k 1x1

UVB 0.028 0.0016

VIS 0.002 0.0049

The QC parameters are not trended and not scored

No remarks.

Algorithm Saturation

The QC parameter is derived from simple pixel counting and filtering by thresholds.

Efficiency (for flux standard stars only)

Signal-to-noise ratio | Saturation | Efficiency (for flux standard stars only) | Response function (for flux standard stars only)

FITS key QC1 database: table, name definition class* HC_plot** more docu

QC.EFF1 xshooter_fluxeff..eff1 mean efficiency in first wavelength region HC [docuSys coming]

QC.EFF2 xshooter_fluxeff..eff2 mean efficiency in second wavelength region HC [docuSys coming]

QC.EFF3 xshooter_fluxeff..eff3 mean efficiency in third wavelength region HC [docuSys coming]

QC.PEAK_EFF_1 xshooter_fluxeff..peak_eff1 HC [docuSys coming]

QC.MED_EFF_1 xshooter_fluxeff..med_eff1 HC [docuSys coming]

*Class: KPI - instrument performance; HC - instrument health; CAL - calibration quality; ENG - engineering parameter
**There might be more than one.

Scoring&thresholds Efficiency (for flux standard stars only)

eff1, eff2, eff3 gives the average efficiency over the i'th wavelength range specified by start and end, see e.g. upper left box of EFF_VIS
peak_eff_i gives the maximum efficiency in order i.
med_eff_i gives the median efficiency in order i. Values of all orders are combined to build a chromatic efficiency plot, see e.g. lower left box of chromatic efficiency UVB

The QC parameters are not scored

Algorithm Efficiency (for flux standard stars only)

The processing of flux standard stars are been improved since 2013-06. The improvements include:

Spectrophotometric reference spectra are based on model atmospheres
For the VIS and NIR-arm the observed flux standard star spectrum is corrected for telluric absorption to improve the stability and quality of the spline fit.
The smoothing of the ratio spectrum (the response function) is no longer performed by a masking technique but by a grid point technique.

For each spectral order the reference spectrum is fit to the observed spectrum. The intensity scaling factor is the efficiency.

Response function (for flux standard stars only)

Signal-to-noise ratio | Saturation | Efficiency (for flux standard stars only) | Response function (for flux standard stars only)

While the efficiency is a technical parameter being independent on the flat field lamp, (the efficiency as a function of wavelength shows the blaze function), the response function is a flat fielded product being dependent on the lamp spectrum. The response function is parameterized and monitored in the same way as the flat lamp spectra. The flux standard star acquired in 2015-10-24 and the resulting response function serves as a reference response function to which the response function being derived from the nearly daily acquired flux standard star is compared with.

Example QC reports on the response function for LTT3128 acquired on 2013-06-16 are given below:

raw response function, cleaned response function (spline fit) and spline fit grid points calibrated flux standard star spectrum, reference model spectrum overplotted calibrated flux standard star spectrum, reference model spectrum overplotted, zoomed version

UVB

VIS

NIR

Comments on the Figures: UVB:

The wiggles in the UVB arm at 330 nm are telluric ozone, OIII features not covered by the model.
The prominent absorption at 570 nm is an instrumental effect.
The discontinuity of the response at ~370 nm comes from the flat field, which is composed of two lamps covering different orders of the echelle spectrum.

NIR:

The response function given in the plot is corrected for telluric absorption.
The background in the K-band is high exceeding the 42000 ADU limit of the NIR-detector. Pixel with extrapolated flux (> 42000 ADU) are marked with the tag 'saturated'

UVB-arm, D2-lamp calibrated spectral region, current response function (black), reference response function from 2015 (red) and current response function aligned to the reference response function with a linear least square fit (green).

UVB-arm, QTH-lamp calibrated spectral region, current response function (black), reference response function from 2015 (red) and current response function aligned to the reference response function with a linear least square fit (green).

VIS-arm current response function (black), reference response function from 2015 (red) and current response function aligned to the reference response function with a linear least square fit (green). The intensity has been scaled by 2% (a=1.021) and the gradient has been modified by b=4.47e-5. Top box: the full wavelength range of the arm. Bottom: four zoomed in spectral regions.

NIR-arm current response function (black), reference response function from 2015 (red) and current response function aligned to the reference response function with a linear least square fit (green).

FITS key QC1 database: table, name definition class* HC_plot** more docu

[calculated by QC procedure] xshooter_fluxeff..response_spec_a current/reference spectrum ratio HC [docuSys coming]

[calculated by QC procedure] xshooter_fluxeff..response_spec_b current/reference spectrum slope HC [docuSys coming]

[calculated by QC procedure] xshooter_fluxeff..response_spec_da error in cur/ref spectrum ratio HC [docuSys coming]

[calculated by QC procedure] xshooter_fluxeff..response_spec_db error in cur/ref spectrum slope HC [docuSys coming]

*Class: KPI - instrument performance; HC - instrument health; CAL - calibration quality; ENG - engineering parameter
**There might be more than one.

Scoring&thresholds Response function (for flux standard stars only)

a is the overall intensity scaling parameter and compensates for the daily altering atmospheric conditions and potential aging of the instruments throughput

b is the first order chromatic deviation from the reference response function, a spectral gradient.

The QC parameters are not scored