The 2 stream-exact single scattering (2S-ESS) radiative transfer model. (January 2023)

• Record Type: Journal Article
• Authors: Natraj, V.; Spurr, R.; Gao, A.; Le, T.; Zeng, Z.C.; Fan, S.; Yung, Y.L.
• Is Part Of: Journal of quantitative spectroscopy & radiative transfer. Volume 295(2023)
• Publication Date: 2023-01
• Subjects: Two stream -- Exact single scattering -- Radiative transfer -- Spherical geometry; Spectrum analysis -- Periodicals; Radiation -- Periodicals; Analyse spectrale -- Périodiques; Rayonnement -- Périodiques; Radiation; Spectrum analysis; Periodicals; 543.0858
• DOI: 10.1016/j.jqsrt.2022.108416
• Languages: English
• ISSNs: 0022-4073

Abstract

Highlights: Two-stream approximation is not accurate enough for calculating spectral radiances. 2S-ESS model performs exact single but two-stream multiple scattering calculation. 2S-ESS approximation introduces less than a few percent error in most situations. 2S-ESS model provides three orders of magnitude increase in computational speed. Code is publicly available along with documentation and test cases to assist users. Abstract: The plane-parallel two-stream approximation is a popular radiative transfer approach for the calculation of fluxes and heating rates. However, it is typically not accurate enough for remote sensing applications involving the analysis of hyperspectral radiances. We present the 2 stream-exact single scattering (2S-ESS) radiative transfer model, which performs an exact calculation of single scattering in a spherically curved medium using an accurate treatment of the phase function and curved ray-tracing of the solar and line-of-sight paths, while approximating multiple scattering with the plane-parallel two-stream approach. The 2S-ESS model has three important features. First, it can be deployed for calculations in vertically inhomogeneous atmospheres. Second, the sphericity capability makes it applicable to large solar and/or viewing zenith angle scenarios such as those encountered close to sunrise or sunset. Third, it is fully linearized: in addition to generating radiances, the model can also compute Jacobians analytically with respect to any atmospheric or surface property (e.g., trace gases, aerosols and surface reflectance). These features of the model are especially useful for remote sensing retrieval applications. We examine the accuracy of this model for homogeneous slab and inhomogeneous multi-layer scenarios. The results show that this methodology introduces less than a few percent error in most situations, with the exception of heavy aerosol or cloud loading events, while providing three orders of magnitude improvement in computational efficiency. The code is publicly available along with documentation and test cases to assist the user.