QC PLOTS
	LAST	HISTORY
frame zeropoints
QC1 database (advanced users): browse | plot
night zeropoints
QC1 database (advanced users): browse | plot
Click on CURRENT to see the current trending (Health Check).    Click on HISTORY to see the historical evolution of the trending.

Warning: Do not use these zeropoints for science data. They are reduced assuming standard extinction coefficients and colour terms, which may not be appropriate for an individual night! These were hardcoded in the old pipeline (for data observed until 2008-03-31) and are tabulated in configuration files for the new pipeline (for data observed since 2008-04-01).

Photometric zeropoints have been calculated starting from exposures of standard stars taken from the Landolt catalogue (Landolt 1992, AJ 104, 340). There are two kinds of zeropoints - per frame and per night. Average zeropoints can be found here. The raw data have been:

• de-biassed,
• flat-fielded with a normalized twilight sky flat.

The twilight images of FORS1, which are used to create master sky flats, show evidence for a feature, whose position depends on the rotator angle (see overview). This limits the achievable photometric accuracy to about 5%. For more information please see Moehler et al. (2010, PASP 122, 93).

In order to select stars observed during photometric nights we check for each night the flux variation observed by the Line Of Sight Sky Absorption Monitor. If the variation exceeds 0.01+1.5*sigma, where sigma is the rms error of the flux variation, the night is flagged as probably non-photometric. This classification is then verified by checking the plot.

QC1 parameters

• zeropoint per frame (QC1 database table fors1_zp_frame, column zp_frame)
  The SExtractor (Source Extractor) software (Bertin & Arnouts, 1996, A&A 117, 393) is run on the reduced images, to detect standard stars and to extract their fluxes. A fixed aperture of 10" radius is used and flux from neighbouring stars within this aperture is corrected. The fluxes of the unsaturated stars are then converted from ADU per sec into electrons per sec using the FITS keyword "HIERARCH ESO DET OUT1 CONAD" and corrected for atmospheric extinction and colour terms using tabulated coefficients. For the zero points per frame the optimally weighted average of the zeropoints from the individual unsaturated stars is determined. The optimal weights are determined in such a way, that the error on the frame zeropoint is minimised. Individual zeropoints deviating more than either 0.3 mag or 5 σ from the mean zeropoint are rejected as outliers, and the mean is re-determined.
  Washing or re-coating of mirrors affects the zeropoints. The dates of such interventions can be found here.
• error of the zeropoint (QC1 database table fors1_zp_frame, column sig_zp_frame)
  This is the combination of the rms error of the individual zeropoints and the accumulated errors from the frames (raw frame, master bias, master twilight flat).

Trending

For the trending we select only zeropoints with errors of less than 0.15 mag. We distinguish three groups of zeropoints

• data from frames with more than one standard star and errors below 0.1 mag, observed during photometric nights (red filled dots)
• data from frames with more than one standard star observed during non-photometric nights (black filled dots)
• data from just one standard star (open lozenges)
  

In addition we trend the error of the frame zeropoints to check if outlying zeropoints have larger than usual errors.

History

• Washing or re-coating of mirrors affects the zeropoints, as does the move from one telescope to another. The dates of such interventions can be found here.
• For the old FORS1 CCD (in operation until 2007-03-25) the conversion factors were read from the FITS keywords "HIERARCH ESO DET OUTi CONAD" with i = 1...4 and averaged over the four CCD ports.
• Before 1999-12-11, a different method had been used, the adaptive aperture method using "first moments". This gives magnitudes which are systematically lower by about 0.06 mags or more.
• Until March 2006, the instrumental magnitudes were estimated with the MIDAS intergrate/aperture command instead of using the SExtractor magnitudes from the pipeline, which were unrealiable due to an incorrect setting of the SExtractor MASK parameter. The instrumental magnitudes were corrected for extinction and colour using coefficients which were read from the QC1 Database and updated after every period.
• Until 2008-03-31, the old imaging pipeline was used, which determined the zeropoints as follows:
• zeropoint per frame (QC1 database table fors1_zp_frame, column zp_frame)
  The SExtractor (Source Extractor) software (Bertin & Arnouts, 1996, A&A 117, 393) is run on the reduced images, to detect standard stars and to extract their fluxes. A fixed aperture of 10" radius is used. The fluxes of the unsaturated stars are then converted from ADU per sec into electrons per sec using the FITS keyword "HIERARCH ESO DET OUT1 CONAD" and corrected for atmospheric extinction and colour terms using hardcoded coefficients. For the zero points per frame the average of the zeropoints from the individual unsaturated stars is determined. Zeropoints fainter than this average by more than 0.3 mag are excluded and the average is re-determined.
• error of the zeropoint (QC1 database table fors1_zp_frame, column sig_zp_frame)
  This is the rms error of the frame zeropoint (0 in general means that there was just one standard star in the field)
• Since April 2007 FORS1 has a blue-sensitive CCD mosaic and new set of high-throughput filters (u_HIGH, b_HIGH, v_HIGH, g_HIGH)
• Since April 2008 the new imaging pipeline has been in use. It uses the Stetson catalogue, which does not contain U band data. Therefore U band (U_BESS, u_HIGH) zeropoints are not available anymore. At the same time the standard filter set was changed to R_BESS, I_BESS, b_HIGH, v_HIGH, and g_HIGH.

QC1 parameters

• extinction coefficient (QC1 database table fors1_zp_frame, column ext_coeff)
  Assuming a value for the zeropoint an extinction coefficient is derived. The values assumed for the zeropoints were derived from photometric nights for a whole period (see here for details) and are given here.
• error of the extinction coefficient (QC1 database table fors1_zp_frame, column err_ext_coeff)
  This is the rms error of the extinction coefficient.

QC1 parameters

• zeropoint per night (QC1 database table fors1_zp_night, column zeropoint)
  For the zeropoints per night the individual zeropoints per star, uncorrected for extinction and colour terms, are used after a pre-selection is done: If there are more than 3 stars in a given frame, a linear regression of zeropoint versus colour is obtained. Stars deviating by more than 1 sigma (uncertainty of fit) from this relation are discarded. Then all stars remaining for that night are used to determine the zero point for the night. The night zero points are determined applying extinction coefficients and colour terms which are read from the QC1 Database. These coefficients are updated after every period and may cause jumps of up to 0.1 mag in the night zeropoints, if large changes occur.
• error of the zeropoint (QC1 database table fors1_zp_night, column sigma_zeropoint)
  This is the rms error of the night zeropoint.

Trending

For the trending we select only zeropoints with errors of less than 0.15 mag. We distinguish three groups of zeropoints

• data with more than one standard star (on average over all filters) and errors below 0.1 mag, observed during photometric nights (red filled dots)
• data with more than one standard star (on average over all filters) observed during non-photometric nights (black filled dots)
• data from just one standard star (open lozenges)
  

In addition we trend the error of the night zeropoints to check if outlying zeropoints have larger than usual errors.

History

• Washing or re-coating of mirrors affects the zeropoints, as does the move from one telescope to another. The dates of such interventions can be found here.
• Until March 2006, the instrumental magnitudes were estimated with the MIDAS intergrate/aperture command instead of using the SExtractor magnitudes from the pipeline, which were unrealiable due to an incorrect setting of the SExtractor MASK parameter.
• Since April 2007 FORS1 has a blue-sensitive CCD mosaic and new set of high-throughput filters (u_HIGH, b_HIGH, v_HIGH, g_HIGH)
• Since April 2008 the new imaging pipeline has been in use. It uses the Stetson catalogue, which does not contain U band data. Therefore U band (U_BESS, u_HIGH) zeropoints are not available anymore. At the same time the standard filter set was changed to R_BESS, I_BESS, b_HIGH, v_HIGH, and g_HIGH.