Spectroscopic measurements of the auroral lines 5577, 4278, and 6300 ¿ made at Fort Yukon, Alaska, are used in the model computations of Rees and Luckey (1974) to predict the energy influx and the characteristic energy of an assumed Maxwellian primary electron spectrum for two auroral displays. Simultaneous with the ground observations, electron detectors aboard a sounding rocket directly measured the primary electron spectrum and energy flux on the field lines which contained the auroral light in the E region observed by the ground photometers (magnetically conjugate in the local sense). For the two auroras studied, the in situ particle measurements show that the model (1) correctly predicts changes in spectral parameters. (2) predicts a precipitated energy flux in good agreement with measured values, and (3) assumes a spectral shape (Maxwellian) not typical of the peaked spectra measured above discrete auroras.

One of the rocket flights also carried photometers sensitive to 5577 and 3914 ¿. Every 0.2 s the photometers sampled the auroral light from the E region magnetically conjugate to the rocket, and they have reaffirmed the very close correlation between emission at 3914 ¿ and that at 5577 ¿. Finally, by using the measured electron precipitation and current ionospheric models the emissions at 3914, 4278, and 5577 ¿ are calculated. The model computations closely predict the measured light at 3914 and 4278 ¿. However, the 5577-¿ emission calculated from dissociative recombination of O2+ and direct excitation of atomic oxygen using a measured secondary spectrum accounts for only about one third of the observed emission.