Validation of spectroscopic data in the 1.27 µm spectral region by comparisons with ground-based atmospheric measurements

Abstract

Various spectroscopic data for absorption lines due to the magnetic dipole transitions of the a1Δg−X3Σg−(0−0) band of O2 centered at 1.27 µm are tested by comparison with high-resolution ground-based atmospheric measurements recorded by Fourier Transform Spectrometers at Park Falls and Caltech (USA). This band is of importance for atmospheric remote sensing since it will be used (together with the O2 A-band near 760 nm) by the passive short wave infrared spectrometer onboard the MicroCarb satellite mission (i.e. this band includes the B4 band of MicroCarb, from about 7800 cm−1 to 7912 cm−1) for the determination of surface pressure and atmospheric aerosols. Spectroscopic data of the HITRAN2016 and GEISA2015 databases as well as those from recent laboratory studies are here used in a radiative transfer code to simulate atmospheric transmissions under the conditions of the measurements. Comparisons are made for different solar zenith angles and for the whole B4 spectral range considered by MicroCarb. Spectroscopic data of water vapor are also tested by considering both relatively dry and humid atmospheric conditions. Averaging the “calculated-observed” residuals over numerous recordings made for close values of the solar zenith angle and humidity enables reduction of the uncertainties due to the radiometric noise of the instrument and to the imperfect description of the atmospheric state. This enables the detection of systematic differences in the spectral residuals caused by small changes in spectroscopic data. The results show that the spectroscopic parameters in the HITRAN2016 and GEISA2015 databases lead to large residuals while data of two recent laboratory studies, obtained from spectra measured with the cavity ring-down spectroscopy technique using the speed-dependent Nelkin-Ghatak profile lead to much better agreement with atmospheric measurements. Significant residuals are noted for water vapor absorption lines simulated using parameters provided by both the HITRAN and GEISA databases.