Spectral dependency of optical backscattering by marine particles from satellite remote sensing of the global ocean

Abstract

[1] Knowledge of the relative proportion between small-sized and larger particles in the surface ocean is essential to understand the ocean ecology and biogeochemistry, including particle dynamics and carbon cycling. We show that this information may be assessed qualitatively from satellite observations of ocean color. Such capability is based on the estimation of spectral dependence, g, of particulate backscattering coefficient, b bp , which is sensitive to particle size distribution. Our results obtained from satellite observations of the global ocean are supported by in situ measurements, and they demonstrate a general decrease of the spectral slope g from oligotrophic to eutrophic regimes, although significant regional differences are observed in the relationship between g and the chlorophyll a concentration, Chl. To first approximation, such a decrease in g is expected to be accompanied by an increased role of larger particles. This is consistent with our field data that show relatively high concentrations of submicron particles in very clear oceanic waters. Different seasonal patterns are also observed depending on the oceanic regions. The seasonal amplitude of g is generally higher than that of Chl and b bp in equatorial and tropical regions, and it is much lower at temperate latitudes. These spatio-temporal patterns are interpreted in terms of processes that modify the composition of particulate assemblages and physiology of phytoplankton in response to environmental forcing. The changes in g are clearly related to variations in the mixed layer depth and photosynthetic available radiation.

Generalized ocean color inversion model for retrieving marine inherent optical properties

“…LAS sequentially estimates b bp λ i and a λ i at each wavelength λ i using K d λ i , environmental geometries, and LUTs derived from radiative transfer (Monte Carlo) simulations. By estimating IOPs at each wavelength independently, spectral shape functions, such as S bp , can be calculated dynamically and considered output products [60]. GIOP supports the use of LAS-derived S bp and (tabulated) b bp λ as Fig.…”

A regional comparison of particle size distributions and the power law approximation in oceanic and estuarine surface waters

“…Open ocean waters lack significant terrigenous input and the optical properties tend to covary with the concentration of phytoplankton, the so-called bio-optical assumption of Case 1 waters [Smith and Baker, 1978]. Since Chl concentrations can be related to phytoplankton size [Yentsch and Phinney, 1989;Chisholm, 1992;Loisel et al, 2006], we hypothesized that higher Chl would more likely be associated with regions of larger particles (lower PSD slope) and lower Chl with regions of smaller particles (higher PSD slope) [Sullivan et al, 2005]. From the stations sampled in the North Atlantic, a significant inverse relationship was found between measured PSD slopes and Chl (Figure 6f).…”

“…This study is restricted to surface PSD measurements (<5 m) to facilitate comparisons between shallow estuarine and deep water oligotrophic stations and to maximize relevance to remote sensing applications. Analysis of small particles (micron to submicron), while potentially significant for optical scattering [Stramski and Kiefer, 1991;Loisel et al, 2006], is beyond the methods employed in this research. By examining particle size distributions in a number of different environments, this study seeks to determine the applicability and limitations of the power law approximation and assess the distributions in a wide variety of natural waters.…”

“…For natural waters, PSD slopes generally vary from 3 to 5 with most values between 3.5 and 4 [Jonasz, 1983] and can be up to 7 for very small submicron particles [Loisel et al, 2006]. For biooptical applications, the PSD slope is often inferred from the spectral slope of the particulate beam attenuation coefficient Twardowski et al, 2001;Sullivan et al, 2005].…”

Optical techniques for remote and in-situ characterization of particles pertinent to GEOTRACES

“…These algorithms, most often, obtain the combined absorption of non-phytoplankton and dissolved material, phytoplankton absorption, the associated chlorophyll concentration, and the backscattering coefficient (see IOCCG, 2006;Werdell et al, 2013). Semi-analytical algorithms provide information regarding size (Loisel et al, 2006;Kostadinov et al, 2009;Berwin et al, 2011) and phytoplankton composition as well (Kostadinov et al, 2010).…”