The observing strategy for the Pilot Survey has been dictated mainly by the present capabilities of the telescope control system. The two major constraints are the absence of an automatic recovery from loss of guide star and of any "batch observation system" (such as p2pp+bob at the NTT). In this case it was decided to adopt a the simple solution of a point and dither strategy for the observations.
Each dithered pointing consists of 5 frames of 4 minutes exposure time for each. The sequence of offsets has been carefully determined in order to minimize the time lost in the settings of the telescope before taking each image. The main constraints was having a dithering sequence with the largest possible offsets between images (to improove the astrometric and photometric accuracy) while keeping the guide star on the tracker CCD. In fact, acquisition of a guide star takes approximately 2 minutes which would be repeated 5 times. It would decrease the efficiency of the survey considerably. The sequence of taking 5 images in a pointing lasted, on average, 28 minutes, while the total integration time was 20 minutes, which resulted in an efficiency of 70%. In this way, it is possible to cover one square degree (4 pointings) per band in 2 hours of observing time.
The dithering scheme adopted is presented in the following figure (red dots). The 5 positions cover a box of 3 by 6 arcminutes. The box of the figure outines approximately the tracker CCD. The blue stripes show the gaps between the CCDs.
A full patch can be covered by 24 pointings (6 columns and 4 rows). Each pointing has an overlap of 2 arcmin on each side with the adjacent pointings. The following figure shows the pointings coverage for patch C (red lines). The green dashed line shows the area of patch C completely covered on the I-band image in the EIS-Wide Survey. The top-left pointing we shows a schematic view of a single image of the WFI.