There are three particular `features' of the WHIRCAM array, that differentiates it from modern optical CCDs in terms of observing procedure. These are residual imaging where a recent high count in any pixel causes a larger count than normal in a subsequent exposure, a high DARK current, and a large number of hot pixels, some of which are unstable and come and go on very short timescales (seconds). The residual imaging effect means that observing a bright object will result in a `ghost' in the subsequent image, and the high DARK current means that if the array is idle for a time, most of the pixels will have a high count, so that a subsequent exposure will have a spurious higher count level. Hence each exposure sequence should be preceded by at least two DARKs, to ensure that the pixel response returns to nominal for the OBJECT exposures. And the large number of hot pixels means that it is essential that each of these OBJECT exposures consists of at least three (preferably more) dithered images, such that median or other suitable filtering can be employed to remove the hot pixels. This applies to standards as well as objects! Dithering also enables accurate sky subtraction for point sources, removes cosmic rays, ensures good temporal and spatial correlation between object and sky, and removes detector dark current automatically. All at no additional cost in observing time.
Fortunately this observing sequence is easily achieved by the use of
EXEC files (see § 3.5.1) that specify a dithered pattern
of exposures per filter,
preceded by two DARKs. Make sure that the dither pattern is not exactly
square, otherwise the bad column ( 
# 55, see Figure 2)
will create a residual series of streaks in the final montage.
You will also need to obtain good estimates of the sky, which is very much
brighter in the IR compared to the optical. For example, at K, the sky
is typically 12 mag/arcsec², compared to 22 mag/arcsec² in V.
This means that most broad band
imaging will be sky limited, and the maximum exposure times to avoid
saturation will be determined by the sky brightness in most observations.
Also, if your aim is to measure surface brightnesses and your source is
extended, or the background is not uniform (eg a gradient
on a galactic field) and covers a large fraction of the WHIRCAM FOV,
then you'll need to take offset skies (ie offset the telescope
to a patch of true sky), of comparable exposure time, similar airmass and
UT as your object exposures. Such offset sky exposures should also be
taken in a dithered pattern, to remove faint stars, hot pixels
and the occasional cosmic ray.