Temperature dependence of the reaction N 2 (A 3 ∑ u + )+O in the terrestrial thermosphere

Abstract

Previous models for dayglow and auroral emissions of the N 2 (A 3 ∑ u + →X 1 ∑ g + ) Vegard‐Kaplan (VK) bands and O( 1 S) lines, when based on laboratory rate coefficients, disagree with observations. The problem has two parts: the overall rate of N 2 (A)+O and the state‐specific yield of O( 1 S). Resolving these discrepancies should yield more accurate determinations of atomic oxygen density by remote sensing of the 2972 and 5577 Å lines. To solve the problem, the sources and sinks Of O( 1 S) are considered using dayglow observations from 105 to 315 km and a numerical model. Line and band intensities are extracted from the data using a multiple regression fit to synthetic spectra. A photoelectron and photochemical model is used to analyze the resulting vertical emission profiles. N 2 Second Positive (2P) altitude profiles indicate that photoelectron excitation of the N 2 triplet system is modeled with an absolute uncertainty of ±23%. The VK/2P intensity ratio suggests that laboratory rate coefficients for the reaction N 2 (A; v '=0,1,2) + O should be increased by a factor of 1.74 to 2.34. However, the laboratory rates were measured at room temperature. When the effect of high thermospheric temperatures on collision frequency is accounted for, the rate coefficients for v ' =0, 1, and 2 are found to be (3.4±0.8)×10 −11 ( T /298) 1/2 , (5.6±1.3)×10 −11 ( T /298) 1/2 , and (4.8±1.2)×10 −11 ( T /298) 1/2 cm 3 S −1 . At 298 K, the v '=0 and 2 values are within 5% of the laboratory values, but for v ' =1 the value is 40% larger than the laboratory value. The effective quantum yield of O( 1 S) by N 2 (A)+O is found to be 0.47+0.17. The observations support a photoelectron cross section for O( 1 S) that is consistent with laboratory measurements, but about 2.0 times larger than theoretical calculations.