Last night, it was cloudy and I made a comparison between the dome and the sky flats. First, I'll analyze dome flats made with the low-intensity flatfield lamps. The image below, ratio_low.ps , is the normalized version of intensity of (dome/sky). When I took dome flats, I tried to point the telescope at the white spot. As you can see, the lower left corner of the dome flat is brighter, and the upper left corner is dimmer, than the rest.
Next, I analyzed dome flats taken with the high-intensity flatfield lamps. The high intensity dome flat shows a slightly different pattern, see ratio_high.ps , shown below. These may suggest that the directions of the lamp are slightly incorrect. (Of course the telescope pointing may be incorrect.)
Using both the (low intensity) dome and sky flats, I processed one image along the galactic disk (NGC6791 field) and ran aperture photometry. I selected clean stars without nearby star contamination. I compared the aperture magnitude between the two and then produced the smoothed map Del_mag.ps (see figure below). (Circles in the map are artifacts from the smoothing process.) In the figure, the unit of the scale bar is given in milli-mag and the map is for AP(sky) - AP(dome). The difference could be as large as 0.015 mag, but it is smaller than that I expected.
