Common Retrieval of Aerosol Properties for Imaging Satellite Sensors

Abstract

We developed a common algorithm to retrieve aerosol properties, such as aerosol optical thickness, single-scattering albedo, and Ångström exponent for various satellite sensors over land and ocean. The three main features of this algorithm are: (1) automatic selection of the optimum channels for aerosol retrieval by introducing a weight for each channel to the object function, (2) setting common candidate aerosol models over land and ocean, and (3) preparing lookup tables for every 1 nm in the range of 300 to 2500 nm in the wavelength and weighting the radiance using the response function for each sensor. This method was applied to the Advanced Himawari Imager (AHI) on board the Japan Meteorological Agency's geostationary satellite Himawari-8, and the results depicted a continuous estimate of aerosol optical thickness over land and ocean. Furthermore, the aerosol optical thickness estimated using our algorithm was generally consistent with the products of the Moderate Resolution Imaging Spectroradiometer (MODIS) and Aerosol Robotic Network (AERONET). In addition, we applied our algorithm to MODIS on board the Aqua satellite and compared the retrieval results to the results obtained from AHI. The comparisons of the aerosol optical thickness, retrieved from different sensors with the different viewing angles onboard the geostationary and polar-orbiting satellites, suggest an underestimation of aerosol optical thickness at the backscattering direction (or overestimated in other directions). The retrieval of aerosol properties using a common algorithm allows in identifying a weakness in the algorithm, such as the assumptions in the aerosol model (e.g., sphericity or size distribution).

Evaluating Aerosol Optical Depth From Himawari‐8 With Sun Photometer Network

“…Given that the surface term was small over ocean, except Sun glitter, and atmospheric path reflectance can be calculated almost precisely, the overestimated AHI ocean AOD500 is likely to be induced by the underestimation of aerosol scattering phase function. The coarse aerosol model used in AHI aerosol algorithm was external mixture of the pure marine aerosol and a dust model (Yoshida et al, ). Priori estimate of dust mixing ratio is zero over ocean, which means that the coarse aerosol model consists of spherical marine aerosol.…”

“…Priori estimate of dust mixing ratio is zero over ocean, which means that the coarse aerosol model consists of spherical marine aerosol. The selected dust mixing ratio is largely determined by priori estimates, especially over ocean, where the observed reflectance is not as sensitive to imaginary part of complex refraction index (Yoshida et al, ). Phase function depends on particle shape model (spherical or nonspherical), but the realistic shape model is different to determine.…”

“…Moreover, a nonspherical model for the coarse dust model was applied in AHI retrievals and the dust nonspherical parameters were based on a yellow dust model ( Nakajima et al, ; Tanaka et al, ). Phase function of spherical model (fine aerosol) is higher than that of nonspherical model at the backscattering direction ( Yoshida et al, ). At Baengnyeong site, a large proportion of scattering angle (~65%) ranges from 100° to 130°, which is tend to retrieve relatively large AOD500 in Figure e.…”

“…The minimum reflectance method is affected by the surface bidirectional reflectance distribution function (Hsu et al, ). In order to reduce the effect of bidirectional reflectance distribution function, surface reflectance was obtained for each pixel where AHI sensor geometries are identical and every hour when the solar geometries are almost identical for a month (Yoshida et al, ). Therefore, the solar zenith angle has less influence on AHI aerosol retrieval, which can be supported by the Figure c.…”

Hourly Aerosol Assimilation of Himawari‐8 AOT Using the Four‐Dimensional Local Ensemble Transform Kalman Filter

“…Himawari‐8, a next‐generation geostationary meteorological satellite, started its official operation on 7 July 2015, is located at 140.7 ° E over the equator and covers the Asia‐Oceania regions from 80 ° E to 160 ° W between 60 ° S to 60 ° N (Bessho et al, ; Kurihara et al, ). The highly improved AHI onboard Himawari‐8 has 16 observation bands and includes observational bands in the visible and near‐infrared wavelengths, is sensitive to aerosol scattering and absorption, and provides full‐disk images of aerosol optical properties, including the AOTs and Ångström Exponent, at 10‐min intervals (Fukuda et al, ; Higurashi & Nakajima, , ; Yoshida et al, ). The Himawari‐8 aerosol optical properties are freely available at the website of the Japan Aerospace Exploration Agency (JAXA) Himawari Monitor (http://www.eorc.jaxa.jp/ptree/index.html).…”