mirror sites: PL (internal link) HQ     [?]

Quality Control and
Data Processing


CAL | HC | refs | QC
QUALITY CONTROL
    HOME
GIRAFFE QC
Trending & QC1
   wavelength solution
Pipeline
GIRAFFE ...
QC links:
GIRAFFE: Wavelength solution and resolution
Resolution | Residual RMS | Number of lines

 
HC PLOTS
Resolution:
Residual RMS:
Number of lines:
QC1 database (advanced users): browse | plot

The quality of the dispersion solution depends on the quality of the arclamp exposure and of the pipeline recipe. The arclamp exposure is quality-controlled in the arclamp plot. The results of the pipeline recipe are monitored here, in terms of three parameters:

  • resolution,
  • residual rms,
  • number of detected emission lines.

Resolution
Resolution | Residual RMS | Number of lines

The resolution of the GIRAFFE gratings is monitored by measuring the width of the identified and accepted arc emission lines in arclamp calibration data. The QC1 parameter mean_resol measures the fibre-averaged width of the fitted emission lines.

QC1_parameters

FITS key QC1 database: table, name definition class* HC_plot** more docu
QC.RESOL.MEAN giraffe_wave..mean_R mean resolving power R = ins_grat_wlen/mean_resolKPI
*Class: KPI - instrument performance; HC - instrument health; CAL - calibration quality; ENG - engineering parameter
**There might be more than one, always check the link 'HC'.

Trending

The upper panel of the trending plot monitors the HC data taken in high resolution (H525.8B, all five fibre systems), the lower panel those for low resolution (L543.1, Medusa1 and 2 only).

History

Date event
around 2008-01-14 Low flux observed, due to bad positioning of the calibration mirror located inside of the OzPoz calibration unit
2010-07-13...24 problems with gripper, calibration no flux
2015-02 Focus of the gratings (in particular of the HR grating) re-aligned. See the FULL plots for residual RMS | number of lines | resolution and Science Newsletter
2015-11-27 strong earthquake, affecting the grating alignment of GIRAFFE (all slit systems) and the quality of the reference dispersion solutions (Medusa2 only). New static calibrations available after complete set of calibrations acquired on 2015-12-06.
2016-10-01 As a result of the re-alignment in 2015-02, the scatter in the plots has become much less, and we have reduced the threshold ranges to +/-10% (empirical value). Before it was at +/-15% or even higher, due to the strong scatter.

In February 2015, the gratings (in particular the HR grating) were re-aligned. The alignment had degraded since about 2010. After the intervention, the resolution was improved by roughly 10% and is now back to the values it had initially. See the FULL plots for residual RMS | number of lines | resolution.

See also the article in the Science Newsletter.

Scoring&thresholds Resolution

The thresholds are relatively tight and set to be indicative of a sudden degradation (due to misalignment).

Algorithm Resolution

mean_resol (QC1 parameter QC.RESOL.MEAN) is transformed into the resolving power R=λ/mean_resol.


Residual RMS
Resolution | Residual RMS | Number of lines

The arc lamp data are measured in a pattern involving two health check settings: the high-resolution H525.8B setup, and the low-resolution L543.1 setup. The former is measured in every slit system, the latter only in the two Medusa's. The pattern repeats every week.

QC1_parameters

The QC1 parameter rms_fit (PRO.WSOL.RMS) describes the rms of the residuals of the accepted line positions against the dispersion solution.

FITS key QC1 database: table, name definition class* HC_plot** more docu
PRO.WSOL.RMS giraffe_wave..rms_fit Average RMS [pxl] of fitted line positions KPI
*Class: KPI - instrument performance; HC - instrument health; CAL - calibration quality; ENG - engineering parameter
**There might be more than one, always check the link 'HC'.

Trending

The trending plot monitors the HC data taken in high resolution (H525.8B, upper panel) and low resolution (L543.1, Medusa1 and 2 only, lower panel).

History

Date event
around 2008-01-14 Low flux observed, due to bad positioning of the calibration mirror located inside of the OzPoz calibration unit
2010-07-13...24 problems with gripper, calibration no flux
2015-02 re-alignment of the focus of the gratings (in particular of the HR grating). The rms_fit values are back to their initial low values. See the FULL plots for residual RMS.
2015-11-27 strong earthquake, affecting the grating alignment of GIRAFFE (all slit systems) and the quality of the reference dispersion solutions (Medusa2 only). New static calibrations available after complete set of calibrations acquired on 2015-12-06.

Scoring&thresholds Residual RMS

The thresholds are set empirically, following the observed behaviour and taking into account two effects: a) the intrinsic scatter of the parameter due to slight changes of the spectral format (thermal motions of the gratings), and b) the longer-term evolution of the parameter due to degrading grating adjustment.

Algorithm Residual RMS

- identify and accept the emission lines from the arclamp per fibre;
- measure their position;
- fit a polynomial;
- calculate the residuals against this fit.


Number of lines
Resolution | Residual RMS | Number of lines

This parameter is derived by the pipeline from the same set of Health Check data as the residual RMS: arclamp file in high resolution (H525.8B) and low resolution (L543.1, Medusa1 and 2 only).

QC1_parameters

FITS key QC1 database: table, name definition class* HC_plot** more docu
PRO.WSOL.NLINES giraffe_wave..nlin_acc average number of accepted emission lines per fibre HC
*Class: KPI - instrument performance; HC - instrument health; CAL - calibration quality; ENG - engineering parameter
**There might be more than one, always check the link 'HC'.

Trending

The upper panel of the trending plot shows the number of accepted lines for the HC settings for high resolution:
Boxes 1-5: Number of accepted lines, high-resolution setting H525.8B.

Plots 6-7 show the results for the HC settings for the low resolution grating:

Boxes 6,7: Number of accepted lines, low-resolution setting L543.3 and Medusa1/2.

For the other settings, this parameter varies strongly. Therefore it is not possible to provide a trending for each individual setup, but there are three plots which monitor the chromatic dependence of nlin_acc:

Boxes 8-10: Number of accepted lines vs. central wavelength; high-resolution settings (all Medusa1/2); high-resolution settings (all IFU1/2 and Argus); low-resolution settings (all Medusa1/2).

History

Date event
around 2007-10-01 arclamp flux strongly decreased due to gripper problem; this resulted in a loss of some of the weaker emission lines, nlin_acc dropped.
around 2008-01-14 Low flux observed, due to bad positioning of the calibration mirror located inside of the OzPoz calibration unit
2010-07-13...24 problems with gripper, calibration no flux
2015-02 re-alignment of the focus of the gratings (in particular of the HR grating). The N_lines values are back to their initial high values. See the FULL plots for number of lines.
2015-11-27 strong earthquake, affecting the grating alignment of GIRAFFE (all slit systems) and the quality of the reference dispersion solutions (Medusa2 only). New static calibrations available after complete set of calibrations acquired on 2015-12-06.
2016-10-01 >As a result of the re-alignment in 2015-02, the scatter in the plots has become much less, and we have reduced the threshold ranges to +/-10% (empirical value). Before it was at +/-15% or even higher, due to the strong scatter.

Scoring&thresholds Number of lines

The thresholds are set relatively tightly, around the current mean value. The QC1 parameter is sensitive to the proper alignment. When the lower threshold is violated this indicates clearly a degradation of the effective spectral resolution. This is clearly visible by comparison to the resolution and rms_fit plots.

Algorithm Number of lines

Lines are identified using a line catalogue, using a noise threshold for weak lines and rejecting saturated lines. Lines not passing a blending criterion are rejected, all others accepted.


Send comments to <qc_flames@eso.org> powered by QC [webCMS v1.0]