[ ESO ]
UVES Quality Control:
Formt_Check

QUALITY CONTROL
HOME
UVES QC
Trending & QC1
   ECH - Grating Position
Pipeline
Data Packages
Data Management
QC links:
uves_fmtchk :: Physical-Model-Prediction-Results | Stability-Check-Results
 
QC PLOTS
  CURRENT HISTORY
Daily Health Checks:    
ECH FMTCHK GratPos PhysM BLUE 346 2x3 DHC
ECH FMTCHK GratPos PhysM BLUE 437 2x3 DHC
ECH FMTCHK GratPos PhysM REDU 580 2x3 DHC
ECH FMTCHK GratPos PhysM REDU 860 2x3 DHC
ECH FMTCHK GratPos PhysM REDL 580 2x3 DHC
ECH FMTCHK GratPos PhysM REDL 860 2x3 DHC
ECH FMTCHK Thermal BLUE DHC
ECH FMTCHK Thermal REDU DHC
ECH FMTCHK Thermal REDL DHC
QC1 uves_fmtchk database (advanced users): browse | plot
   Click on CURRENT to see the last trending (Health Check).
   Click on HISTORY to see the historical evolution of the trending.


top Introduction

This tutorial provides information for the ECH - Grating Position Health Check trending.
The ECH - Grating Position trending is based on the measurement of QC1 parameters of FORMAT CHECK calibration data obtained with dedicated Health Check OBs executed daily. These OBs obtain four sets of FORMAT CHECK data, 2 sets in BLUE mode and 2 sets in RED mode, each set comprising a single RAW frame. Each set is acquired in a different wavelength setting covering the two arms and the four cross dispersers.

top Physical-Model-Prediction-Results

QC1 parameters

  • average temperature of camera [C] (QC1 database table uves_fmtchk, column ins_temp1_mean, product keyword INS.TEMPi.MEAN)
    Description: Mean Temperature during the exposure.
  • difference of X positions to physical model (QC1 database table uves_fmtchk, column mean_dx, product keyword QC.MODEL.DIFFXAVG)
    Description: The physical model used in Formatcheck step allows to measure the difference between model predicted and detected line positions. This is useful to verify precision of physical model prediction and instrument spectral format stability.
  • difference of Y positions to physical model (QC1 database table uves_fmtchk, column mean_dy, product keyword QC.MODEL.DIFFYAVG)
    Description: The physical model used in Formatcheck step allows to measure the difference between model predicted and detected line positions. This is useful to verify precision of physical model prediction and instrument spectral format stability.
  • number of lines selected (QC1 database table uves_fmtchk, column nlin_sel, product keyword QC.MODEL.NLINSEL)
    Description: The physical model used in the Formatcheck step allows to predict line positions on the formatcheck frame. The selected number of lines is the subset of the total number of lines which fall in the physical model plot limits.

 

top Stability-Check-Results

QC1 parameters

  • median of difference in X to reference frame (QC1 database table uves_fmtchk, column median_shftx, product keyword QC.SHFTXMED)
    Description: mmeasured median shift in x from stability check
  • median of difference in Y to reference frame (QC1 database table uves_fmtchk, column median_shfty, product keyword QC.SHFTYMED)
    Description: measured median shift in y from stability check

 


top FORMATCHECK FRAMES

Modes

Formatcheck frames are measured daily whenever UVES is operational. They are used to monitor the orientation of the four crossdispersers and the two gratings. For each crossdisperser, one standard setting is used:

grating setting
1 DI1 346bl_1x1
2 DI2 437bl_1x1
3 DI1 580re_1x1
4 DI2 860re_1x1

QC1 parameters

Three QC1 parameters are measured:

  • the shifts in X and Y of the spectral format in the monitor frame against the physical model (DX_PM, DY_PM),
  • the shift (cross-correlation) of the monitor frame against a fixed reference frame (DX_RF, DY_RF),
  • the number of lines selected by the pipeline recipe (N_sel).

Generally the reference frame method is more sensitive. The shifts are evaluated as function of time and as function of temperature T which is the main parameter for daily drifts.

Typically 300-1100 lines are selected per CCD for this analysis. This parameter is also available as ASCII data set. The actual values are compared to a reference value which is approximately the maximum possible number of lines found.

These QC1 parameters are plotted for all four settings. The red settings have data for the two CCDs (REDU and REDL). A detailed caption is available on the plots.

Y drifts versus temperature

A compensation for thermal drifts in Y (cross-dispersion) direction is active since November 2000 for all four gratings. Before 2000-11, thermally induced Y shifts have been of the order of 0.5 to 1.0 px/K, see sample plot. Since 2000-11, thermal Y shifts of all gratings are automatically compensated. They are now of the same order as X drifts, i.e. roughly 0.2 px/K or less, see sample plot.

parameter procedure applied typical values* plots**
DX_PM, DY_PM (pixel coordinates) against physical model cross-correlation line table (first guess)
BLUE
 
few pixels
REDU
 
few pixels
REDL
 
few pixels
fmtstab
1
: trending of DX_PM and DY_PM vs. time, for the four standard settings and for all three CCDs
DX_RF, DY_RF (pixel coordinates) against reference frame cross-correlation line table (first guess) same as above 2: trending of DX_RF and DY_RF vs. time, for the four standard settings and for all three CCDs
T (in C) read from FITS header raw 0 - 15 C across the year; few C across months 1, 2 (top row): correlation DX_RF, DY_RF with T
N_SEL line identification plus outlier rejection procedure line table (first guess) between 300 (860nm, REDU CCD) and 1100 (437nm, BLUE CCD) 3: trending of N_sel vs. time, for the four standard settings and for all three CCDs

* by August 2000
** refers to numbering in QC plot

Notes

1. A 5.9 mag earthquake on 2000-05-12 shifted the blue gratings by more than -10 pixels (see DY panels 346B, 437B from Apr-Jun2000). The blue gratings have been readjusted on 2000-07-24.

2. In addition to thermal Y drifts, there are from time to time sudden shifts (e.g. earthquake, readjustments). Also slow non-thermal Y drifts (timescale of months) have been observed.

3. Since 2001-12-26, a thermal compensation in X (dispersion) direction has been activated.

4. The trending of N_sel gives an indication about the quality of the DY correlation. If the Y position of a grating starts drifting away from its nominal zeropoint (or is suddenly pushed away), then usually N_sel is the most sensitive parameter for detecting whether or not this is critical and requires immediate reaction. A good example is the period JUL-SEP 2001 where the N_sel parameter for grating 4, REDU CCD, faded away from its reference value (about 320) to less than 200. This indicated a critical regime at least for the UVES pipeline to find the proper dispersion solution.

5. From 2002-05-29 to 2002-06-17, the positons of the red crossdispersers could not be set with high accuracy. This was caused by high friction and contamination of the red crossdisperser bearing. The crossdispersers have been repaired on 2002-06-17. All observations which have been affected by this problem have been re-scheduled.

Events

A list of events which affect the grating positions is available.

 


 
[ESO][Index][Search][Help][News]