UVES Quality Control: WAVE frames

UVES@KUEYEN:
QC1 values from WAVE frames

HOME

INDEX

HELP

NEWS

QC1 main

QC1 PLOTS RESOLUTION/ DISPERSION

QC1 ASCII DATA

CURRENT PERIOD
	gif	ps
resolution
dispersion

(updated regularly)

COMPLETE SET
Year:
Item:
Format:

COMPLETE SET
Year:
Item:
CCD:		BIN:
WLEN:		SLITW:

QC1 database (expert mode)

WAVE frames

Modes

Parameters are measured for all three CCDs separately (BLUE, REDL, REDU). They are available whenever raw arclamp (type LAMP,WAVE) frames are measured in a UVES standard mode.

Since Period 68 (Oct. 01, 2001) a set of daily arclamp exposures is taken which serves as instrument health check, in the two settings BLUE 437 1x1 and RED 580 1x1 and with slit widths 0.4 and 0.3 arcsec, resp. These are used to monitor the resolving power of UVES.

QC1 parameters: resolving power, dispersion, line intensity

The most important QC1 parameters measured are resolving power R and standard deviation of the dispersion solution, sigma. Others are e.g. minimum and maximum wavelength and number of lines selected.

Resolving power R = lambda/delta lambda is evaluated as median over a Gauss fit to all selected lines. R is compared to reference values as function of slit width. The reference values are lab data as taken from the User's Manual.

The quality of the dispersion solution is measured by the standard deviation of the difference between the found and fitted position, again evaluated for all selected lines. Typically 300 - 1000 lines are selected per CCD.

Since March 2004, also the average intensity of a selected number of ThAr lines is evaluated in order to measure the performance of the arc lamp. The results are plotted together with the efficiencies of the flat field lamps (see here).

parameter	procedure	applied	typical values		plots
			setting	value*
effective resolving power R	Gauss fit to selected lines, median of FWHM	line table	BLUE, 0.5":	70 000	resolution 1: trending of R vs. time 2: all R values from the selected period vs. slit width; plus signs and broken curve are nominal values 3:* daily data for the two QC1 settings (monitoring the highest possible, properly sampled values of R)
			REDU, 0.5":	80 000
			REDL, 0.5":	80 000
sigma of dispersion solution	least-square fit of X position of selected lines, standard deviation of residuals	line table	BLUE_1x1/ BLUE_2x2:	1x1: 2-3 mA 2x2: 5-7 mA	dispersion 1: trending of sigma vs. time (mean values given; U, L: red upper, lower CCD); data are presented for all three CCDs, for 1x1 and 2x2 binning; central wavelength of setting is indicated 2:sigma vs. slit width ("); if applicable, a regression is made 3***: daily data for the two QC1 settings
			REDU_1x1/ REDU_2x2:	1x1: 4-6 mA 2x2: 8-15 mA
			REDL_1x1 REDL_2x2:	1x1: 2-4 mA 2x2: 5-9 mA

* in Oct-Dec 2000
** depending on wavelength and slit width
*** since Oct. 01, 2001

FORMAT

All standard settings have one box, separated for 1x1 and 2x2 binnings. Resolving power values are collected in resolution plot 1 regardless of the associated slit width.

Note that the format of the resolution plot has slightly been changed from Oct. 01, 2001 on. The daily QC1 settings have been added. Data for the standard settings are collected per wavelength value (i.e. no distinction is made for non-dichroic or dichroic modes). Data for the newly added RED 600 1x1 setting are put into the RED 580 box.

Note that due to a wrong selection of lines (bad outliers included), previously quoted values for sigma were too high by a factor of 2-4. These numbers have now been corrected both in the plots and in the ASCII tables.

Beginning with pipeline version 1.3, a stricter rule for accepting lines in settings with 2x2 binning has been applied. Therefore, the number of accepted lines and the standard deviation of the dispersion solution are lower for all wavelength calibrations in 2x2 binning from April 07, 2002 on.