The reference sector method relies on two assumptions:

1. That the tropospheric NO₂ over the reference area is negligible.
2. That the stratospheric NO₂ column is a function of latitude only.

The first assumption is usually fulfilled, as air-borne in-situ measurements in the Pacific region show, at least compared with polluted areas such as Europe. However, because the tropospheric NO₂ is not zero, a low bias is introduced. This, in turn, can be corrected by adding a typical NO₂ background, such as one derived from model calculations, for example.

The second assumption is only a first order approximation based on the considerations discussed in the last section. It often fails at higher latitudes (>60°), and in particular in the presence of the polar vortex in spring, thus introducing an uncertainty in the retrieved values. Typical problems can be seen in the figure, where enhanced tropospheric NO₂ appears to be present just south of the equator over the Atlantic Ocean. This artefact disappears when using limb/nadir matching (see next section).

One approach to further improve the reference sector method is to account for the variation of stratospheric NO₂ with longitude either by using model results or by using other unpolluted regions as well and interpolating in between them.