An algorithm for hyperspectral remote sensing of aerosols: 2. Information content analysis for aerosol parameters and principal components of surface spectra. (May 2017)

• Record Type: Journal Article
• Authors: Hou, Weizhen; Wang, Jun; Xu, Xiaoguang; Reid, Jeffrey S.
• Is Part Of: Journal of quantitative spectroscopy & radiative transfer. Volume 192(2017)
• Page Start: 14
• Page End: 29
• Publication Date: 2017-05
• Subjects: TEMPO -- Geostationary satellite -- Information content analysis -- Degrees of freedom for signal -- Principal component -- Hyperspectral remote sensing -- Common band selection; Spectrum analysis -- Periodicals; Radiation -- Periodicals; Analyse spectrale -- Périodiques; Rayonnement -- Périodiques; Radiation; Spectrum analysis; Periodicals; 543.0858
• DOI: 10.1016/j.jqsrt.2017.01.041
• Languages: English
• ISSNs: 0022-4073

Abstract

Abstract: This paper describes the second part of a series of investigation to develop algorithms for simultaneous retrieval of aerosol parameters and surface reflectance from the future hyperspectral and geostationary satellite sensors such as Tropospheric Emissions: Monitoring of POllution (TEMPO). The information content in these hyperspectral measurements is analyzed for 6 principal components (PCs) of surface spectra and a total of 14 aerosol parameters that describe the columnar aerosol volume V total, fine-mode aerosol volume fraction, and the size distribution and wavelength-dependent index of refraction in both coarse and fine mode aerosols. Forward simulations of atmospheric radiative transfer are conducted for 5 surface types (green vegetation, bare soil, rangeland, concrete and mixed surface case) and a wide range of aerosol mixtures. It is shown that the PCs of surface spectra in the atmospheric window channel could be derived from the top-of-the-atmosphere reflectance in the conditions of low aerosol optical depth (AOD ≤ 0.2 at 550 nm), with a relative error of 1%. With degree freedom for signal analysis and the sequential forward selection method, the common bands for different aerosol mixture types and surface types can be selected for aerosol retrieval. The first 20% of our selected bands accounts for more than 90% of information content for aerosols, and only 4 PCs are needed to reconstruct surface reflectance. However, the information content in these common bands from each TEMPO individual observation is insufficient for the simultaneous retrieval of surface's PC weight coefficients and multiple aerosol parameters (other than V total ). In contrast, with multiple observations for the same location from TEMPO in multiple consecutive days, 1–3 additional aerosol parameters could be retrieved. Consequently, a self-adjustable aerosol retrieval algorithm to account for surface types, AOD conditions, and multiple-consecutive observations is recommended to derive aerosol parameters and surface reflectance simultaneously from TEMPO. Highlights: Aerosol information content analysis for hyperspectral satellite sensor in visible. Common bands exist for aerosol retrieval over a wide range of surface conditions. Feasibility study of retrieving both aerosol & surface from geostationary sensor. This feasibility is shown when multiple measurements from geo sensor are used. Self-adjustable algorithm for aerosol is proposed for geostationary sensor.