The geometric area available at the slit, that depends on which slit is used.
The exposure time.
4.5.1 Intensity variations along the CDS slit
For details, see:
http://orpheus.nascom.nasa.gov/~thompson/slit_inten/slit_intensity.html
There is evidence of an intensity gradient along the slit in VDS data.
Furthermore, the variation appears wavelength dependent.
It's unclear what is causing this effect.
It is probably generated within the instrument optics, either within
the telescope or in the two NIS gratings.
Bill Thompson analysed NISAT_S, NIMCP, and SYNOP_F data
and came up with the following conclusions.
The behavior along the slit falls into two classes. One class behaves
like Mg IX 368, where the intensity increases as one goes from the
bottom of the slit to the top. The other class behaves like O V 630 A, where
the intensity descreases as one goes upward from the bottom of the
slit, and then levels off.
Other lines, such as He I 584 A, can be modelled as a linear
combination of these two behaviors.
Figure 27: [Figures from Bill Thompson]
From top to bottom: NIS-1 Mg IX-like behavior;
NIS-2 Mg IX-like behavior; NIS-1 O V-like behavior.