Comparison of rotational temperature derived from ground-based OH airglow observations with TIMED/SABER to evaluate the Einstein coefficients

Ground-based observations of the OH(9-4, 8-3, 6-2, 5-1, and 3-0) band airglows over Xinglong, China (4024′N, 11735′E) from December 2011 to 2014 are used to calculate rotational temperatures. The temperatures are calculated using five commonly used Einstein coefficient data sets. The kinetic temperature from Thermosphere, Ionosphere, Mesosphere, Energetics and Dynamics/Sounding the Atmosphere by Broadband Emission of Radiation (SABER) is completely independent of the OH rotational temperature. SABER temperatures are weighted vertically by weighting functions calculated for each emitting vibrational state from two SABER OH volume emission rate profiles. By comparing the ground-based OH rotational temperature with SABER's, five Einstein coefficient data sets are evaluated. The results show that temporal variations of the rotational temperatures are well correlated with SABER's; the linear correlation coefficients are higher than 0.72, but the slopes of the fit between the SABER and rotational temperatures are not equal to 1. The rotational temperatures calculated using each set of Einstein coefficients produce a different bias with respect to SABER; these are evaluated over each of the vibrational levels to assess the best match. It is concluded that rotational temperatures determined using any of the available Einstein coefficient data sets have systematic errors. However, of the five sets of coefficients, the rotational temperature derived with Langhoff et al.'s (1986) set is most consistent with SABER. In order to get a set of optimal Einstein coefficients for rotational temperature derivation, we derive their ratios from ground-based OH spectra and SABER temperatures statistically using 3 years of data. The use of a standard set of Einstein coefficients will be beneficial for comparing rotational temperatures observed at different sites.