The period of Dec. 8, 1998 through Jan. 18, 1999, was used for theinstallation, testing, commissioning and science verification (SV) of the Wide Field Imager (WFI) at the MPIA 2.2-m telescope on La Silla.  Thetime was shared between ESO and the Max-Planck-Institut f�r Astronomiein Heidelberg; the equivalent of 4 dark nights was used by the ESO ImagingSurvey (EIS) Team.

The ESO part of the SV data (excluding the EIS observations, which willbe published separately) is now offered to the ESO community. The scientificobjectives and targets were selected from a number of suggestions solicitedfrom astronomers at ESO and the Osservatorio Astronomico di Capodimonte. Please consult the list of available exposures for more details.

Unfortunately, the scientific power of the data is reduced with respectto the scope of the original suggestions because of the required broadband filters only the U and B filters were available at the time of SV.However, the U filter turned out to suffer from a severe red leak (thefilter is currently being repaired), which at large air mass manifestsitself in double images.

The SV database contains only raw data. The frames are recorded in FITSformat with 8 image extensions, namely one for each of the 8 CCD's of the8k x 8k science mosaic. With 2 bytes/pixel, their total file size amountsto almost 142 Mbytes. The FITS headers are entirely unedited and, therefore,suffer from a number of errors, omissions and insufficiencies. The mostimportant ones known to date include:

a) RA and DEC refer to a point 15 arcseconds south and westof the northeastern corner of CCD #55.  However, because of problemswith the implementation of the pointing model (not of the pointing modelitself, which had an rms of about 9 arcseconds up to zenith distances of60 degrees), RA is towards very southern declinations increasingly offby up to +1000 arcseconds (not: time seconds).

b)  All negative declinations appear with a positive sign.

c)  UTC is incorrect but DATE-OBS contains the correctUTC.

d)  In many files only the slot number in the filter storagering is given but not the filter name.  The following cross identificationtable can be used in these cases:

The database comprises only very few bias files. This is a deliberatedecision, because the bias level must be extracted from the horizontaloverscan pixels of each CCD. Because it depends on the exposure level inthe corresponding light sensitive rows, increases roughly logarithmicallyover the first couple of hundred rows, and can jump after every re-startof the control software, dedicated bias frames would be of very littleuse.

Observations of photometric standards (Landolt selected areas) wereoften made before the telescope was focussed in order to save time. Butthis does not affect the validity of the data. A crude photometric calibrationof the B-band data is possible using the following relation:

where m_B is the B magnitude, ADU is the number of analog-digital unitsof the object to be calibrated, and (B-V) is the color of the object.

During the SV run, the 2.2-m telescope suffered from significant astigmatism.This resulted in telescope position-dependent elongations of the pointspread function (PSF) of up to 20%. Within each image, the PSF is very homogeneous across the entire field of view.

The above enumeration of imperfections reads rather sobering. However, although it is as honest as current knowledge permits, it is notthe full truth:  The Wide Field Imager does produce superb data, andalso the SV observations have a high scientific value.   As aquick, partial preview, have a look at the ESO Public Image Archive which also contains pictures obtained during the SV run

For more details about properties of the WFI and information missingfrom the FITS headers (e.g., about the CCD characteristics) kindly consultthe usermanual.

Copies of files can be requested from the ESO Science Archive by selectingthem from the ESO archive page. We apologize for the inconvenience of the late releasewhich is mostly due to technical and resource reasons.

The distribution policy is the same as for VLT SV data, i.e. the scienceverification data is available to all archive users with an affiliationin an ESO member state or in Chile or in Portugal. The data will be publiclyavailable worldwide one year after the observation date.

Except for many calibration exposures, the observations were made byDietrich Baade (dbaade@eso.org), whoshould be contacted if further information is needed.