A convenient laboratory calibration method is outlined. Uncertainties introduced by instrumental and atmospheric effects are discussed. The conclusion is that laboratory calibration is to be preferred. The variable atmospheric extinction can introduce great errors in ground-based measurements, and therefore necessitates separate monitoring. The over-all uncertainty of ground-based measurements can be below 20%.
Kjell Henriksen, "Absolute Calibration of Auroral Photometers and Ground-Based Measurements in the 3200–7500-Å Spectral Region," Appl. Opt. 13, 1196-1201 (1974)