Abstract

A quasi-linear retrieval was developed to profile moderately thin atmospheres using a high-resolution O2 A-band spectrometer. The retrieval is explicitly linear with respect to single scattering; the multiple-scattering contribution is treated as a perturbation. The properties of the linear inversion, examined using singular value decomposition of the kernel function, demonstrate the impacts of instrument specifications, such as resolution, out-of-band rejection, and signal-to-noise ratio, on information content. A system with 0.5 cm−1 resolution, signal-to-noise ratio of 100:1, and out-of-band floor of 10−3 has four independent pieces of information. A fast radiative transfer model was developed to compute the multiple-scattering perturbation, in which multiple scattering is calculated at 16 different O2 absorption depths to synthesize the O2 A band. The linear system is then solved using Tikhonov's regularization with inequality constraints. Tests with synthetic data, including noise, of O2 A-band retrievals illustrate that this algorithm is accurate and fast for retrieving aerosol profiles. The errors are less than 10% for the integrated total optical depth for the cases tested. It is shown that instruments with the needed performance are practical.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call