In this study, we compare sprite streamer modeling results with Imager of Sprites and Upper Atmospheric Lightning (ISUAL) spectrophotometric data for several sprite events. The model positive streamers are simulated for two representative magnitudes of the quasi-electrostatic field produced by cloud-to-ground lightning discharges, reflecting conditions at 70 km altitude during the initial stage of sprite formation. The intensity ratio of the second positive band system of N-2 (2PN(2)) to the first negative system of N-2(+) (1NN(2)(+)) is obtained separately from the modeling and the ISUAL measurements. The comparison results indicate that the ratio obtained for the streamer developing in an electric field close to the conventional breakdown threshold field E-k agrees with the ISUAL measurements at the very early stage of the sprite development better than for the streamer developing in a field much lower than E-k. This finding supports the sprite theory proposing that sprites are caused by conventional breakdown of air when the lightning field in the upper atmosphere exceeds the local breakdown threshold field, which has also been supported by a recent study by Hu et al. (2007) comparing modeled lightning fields obtained using measured current moments of causative lightning discharges and video observations of sprites. The fact that the early stage emissions from the sprites under study are well explained by the radiation from streamers in strong fields allows the use of ISUAL data to gain additional information on the poorly known quenching altitude of the N-2(a(1)Pi(g)) state, which is responsible for N-2 Lyman-Birge-Hopfield band system. The results confirm that the 77 km suggested in previous study by Liu and Pasko ( 2005) is a good estimate for the quenching altitude of N-2(a(1)Pi(g)).
