top GENERAL INFORMATION

This page gives an overview of the sky background as derived from observations with VLT instruments. For published measurements of the sky brightness at Paranal see Patat (2009, A&A 481, 575 and references therein). We distinguish below between sky brightness as derived from imaging data and sky spectra.

This monitoring was done by processing science frames. With the rescoping of the QC group processes, science processing was terminated on Sep. 30, 2011, and so was the sky monitoring. Still, the historical information could be useful.

top IMAGING

For FORS1/2 the sky brightness is measured from the reduced SCIENCE images by running a code created by F. Patat (ESO) and described in A&A 401, 797 (2003). Images are scanned to select those best suited for a reliable sky brightness measurement. The image sky background is evaluated automatically and those images affected by a very noisy backgrounds, by enhanced objects crowding, by saturation problems and by reflections or ghost images are filtered out. The measured sky brightness values (in instrumental magnitudes/arcsec sq.) are then calibrated on-the-fly applying the zeropoints, colour terms and extinction coefficients. Applying the extinction coefficients is actually not correct as much of the sky background is created within the atmosphere (see Patat, 2003, sect. 3). FORS1 data cover the time range April 2000 - March 2009, while FORS2 data are determined since March 2007.

The HAWK-I pipeline estimates the sky background (instrumental units only) in the combined images created from the frames provided in the SCIENCE image stack. HAWK-I has provided these measurements since April 2008.

For VIRCAM the pipeline determines the sky background (instrumental units only) via a clipped median.The values have been stored since January 2009.

top SPECTROSCOPY

Most spectroscopic observations require a correction of sky background, either by fitting the sky spectrum (done for most optical observations) or by subtracting an observed sky spectrum. The latter method is prefered in the infra-red wavelength range and consists either of nodding (the target is at different positions along the slit) or offset (besides the position on the target also an offset position containing only sky background is observed) observations. However, not all pipelines keep these sky background products. Below we list those instruments for which sky background spectrum is kept.

Both FORS1/2 and VIMOS provided only fitted sky spectra, i.e. sky background data obtained from the SCIENCE frames under certain assumptions.For FORS1/2 and VIMOS multi-object spectroscopy the sky background is modeled before rectifying the frame. Since this does not work well for curved or slanted slits, for FORS1/2 long slit data (producing curved lines due to the length of the slit) the sky background is determined as the median of the rectified frame. Both approaches assume that at least 50% of the pixels in the exposure contain flux only from the sky background. Both two- and one-dimensional rectifed sky background spectra are stored. The data are not flux-calibrated.

X-Shooter on the other hand provides both observed and fitted sky spectra. Observed sky spectra are derived from OFFSET observations, which alternate between a position on the target and an offset position for the sky background. These sky observations are processed in the same way as the on-target data and provide two-dimensional rectified and merged sky spectra. X-Shooter also provides fitted sky background data for data taken in STARE mode, but as above these might be contaminated by flux from the target in case of bright and/or extended targets. The data are not flux-calibrated.

X-Shooter provides both observed and fitted sky spectra. Observed sky spectra are derived from OFFSET observations, which alternate between a position on the target and an offset position for the sky background. These sky observations are processed in the same way as the on-target data and provide two-dimensional rectified and merged sky spectra. X-Shooter also provides fitted sky background data for data taken in STARE mode, but as above these might be contaminated by flux from the target in case of bright and/or extended targets. Sometimes ther are also problem with the fitting of the sky spectrum in the K band (using the default parameters). The data are not flux-calibrated.

For SINFONI sky spectra are derived also from OFFSET observations. They are processed in the same way as the on-target data and provide cubes for each individual sky frame as well as a median sky cube. The data are not flux-calibrated.

VISIR as the only mid-IR VLT instrument also provides observed sky spectra from OFFSET observations, which are not flux-calibrated.