Retrieval of aerosol optical thickness and surface parameters based on multi-spectral and multi-viewing space-borne measurements. (November 2020)

• Record Type: Journal Article
• Authors: Vountas, Marco; Belinska, Kristina; Rozanov, Vladimir V.; Lelli, Luca; Mei, Linlu; Jafariserajehlou, Soheila; Burrows, John P.
• Is Part Of: Journal of quantitative spectroscopy & radiative transfer. Volume 256(2020)
• Publication Date: 2020-11
• Subjects: Aerosols -- BRDF -- Retrieval -- Remote sensing -- Satellite; Spectrum analysis -- Periodicals; Radiation -- Periodicals; Analyse spectrale -- Périodiques; Rayonnement -- Périodiques; Radiation; Spectrum analysis; Periodicals; 543.0858
• DOI: 10.1016/j.jqsrt.2020.107311
• Languages: English
• ISSNs: 0022-4073

Abstract

Highlights: A novel satellite-based full physical retrieval to derive Aerosol Optical Thickness and Surface parameters over vegetation dominated scenes has been developed. The approach is based on the SCIATRAN and relies on multi-spectral and multi-viewing capabilities of the used satellite instrument. The minimization is done by solving the Quadratic Programming Problem (QPP). The retrieval has been applied to POLDER/PARASOL data and provides very convincing results at native spatial resolution. A first, brief validation with ARM ground based measurements shows good and partly very good agreement (R = 0.84). Abstract: We present a novel approach to derive Aerosol Optical Thickness (AOT) at 0.5 µm and the surface reflectance for five spectral channels at native spatial resolution from the measurements of the Polarization and Directionality of Earth's Reflectances-3 (POLDER) instrument aboard the Polarization & Anisotropy of Reflectances for Atmospheric Sciences coupled with Observations from a Lidar (PARASOL) satellite. POLDER had multi-spectral and multi-viewing capabilities. This publication describes the first step in the design of a high-quality AOT and surface property retrieval algorithm, enabling the global evaluation of future missions providing multi-spectral and multi-viewing space-borne measurements. The developed retrieval approach is based on the radiative transfer and retrieval model SCIATRAN using an analytical linearized retrieval mode. The surface is parametrized according to the Ross-Li model and the aerosol typing was using prescribed types based on the approach by Levy et al. [26]. The minimization using SCIATRAN has been done by solving the quadratic programming problem which minimizes the considered system of equations (typically) based on linear constraints. In this study we constrained the retrieval so far only by assuming that the retrieved AOT at 0.5 µm is larger than 0.01 and smaller than 1.5. Until now, the retrieval is based only upon unpolarized POLDER data and shows promisingly weak dependence on a priori information. As a first validation, the retrieved AOT values have been compared with ground based measurements over Atmospheric Radiation Measurement Climate Research Facilities (ARM) in the Southern Great Plains (SGP). The SCIATRAN retrievals are mostly homogeneous and consistent in the region around the US ARM/SGP measurement stations. The retrieval results agree well with the ARM ground based measurements of AOT. The correlation coefficient was R = 0.84 and part of the remaining differences between ARM measurements and SCIATRAN retrievals can be attributed to unmasked clouds. Histogram based analysis of the AOT values showed reasonable distributions using SCIATRAN. Most distributions based on SCIATRAN AOT retrievals within a radius of 25 km around the stations are significantly broader than those derived from ARM measurements during 30 minutes but are well centered around the ARM AOT distributions.