Effects of microbial particles on oceanic optics: A database of single-particle optical properties

We describe a database of the single-particle optical properties cd marine microbial particles. This database includes representatives from five classes of particles: viruses (VIR), heterotrophic bacteria (BAC), cyanobacteria (CYA), small nanoplanktonic diatoms (DIA), and nanoplanktonic chlorophytes (CHLO). The optical properties of VIR, whose mean size is 0.07 pm, were determined from Mie scattering calculations using reasonable approximations about the size distribution and refractive index of viral particles. The database for BAC, CYA, DIA, and CHL0 was created from laboratory measurements of microbial cultures and modeling of particle optics. BAC are represented by a mixed natural population of bacterial species (-0.55 pm in size), CYA by Synechococcus (clone WH 8103, 1 pm), DIA by Thalassiosira pseudonana (-4 Fm), and CHL0 by Dunaliella tertiolecta (-7.5 pm). The database includes the single-particle optical properties that are useful in radiative transfer modeling: the absorption and scattering cross sections and scattering phase functions. Additionally, the database includes the attenuation cross sections, optical efficiency factors, single-scattering albedos, and backscattering properties. For phytoplankton species, chlorophylland carbon-specific optical coefficients are also available. The optical quantities are generally determined at I-nm intervals in the spectral region from 350 to 750 nm. The scattering phase function is determined at 5-nm intervals in wavelength and 1” intervals in scattering angle. The size distribution and refractive index of the particles are also included. This database, when combined with radiative transfer modeling, provides a powerful approach to advancing our understanding of oceanic optics. Achieving a detailed understanding of the roles played by various types of particles in oceanic radiative transfer is a prerequisite to advancing our knowledge of the marine optical environment and applying optical techniques to the study of the ocean. It is the fundamental processes of absorption and scattering of light by many individual particles and soluble materials present in seawater that determine the magnitudes and variability of the bulk optical properties and light field characteristics in the ocean. To achieve a comprehensive understanding of the marine optical environment and to construct reliable predictive and inverse optical models, we must first understand how each of these various water components absorbs and scatters light, and hence affects underwater radiative transfer. It is well known that there is great variability in the inherent optical properties (IOPs) of seawater in the world’s oceans. This variability induces a correspondingly large variability in oceanic apparent optical properties (AOPs), even after accounting for variability in the AOPs associated with the incident lighting and sea state. The optical properties of Acknowledgments The development of the database was supported by the ONR Environmental Optics Program grants NO00 I 4-88-J12 16, N0001493-l -0134, and N00014-95-l-0491 as well as NASA grant NAGW3574. Use of the database in radiative transfer modeling is being supported by the ONR Environmental Optics Program contract N00014-94-C-0114. We thank A. Morel for comments on the manuscript. 538 seawater are determined in large part by the suspended particulates, and the optical variability can be traced primarily to the biological components of the water, at least in case 1 waters where phytoplankton and co-varying materials are the dominant Icomponents (Morel and Prieur 1977). Most efforts to date have been focused on developing bio-optical relationships :For such waters using a parameterization of seawater composition in terms of chlorophyll concentration alone. It has been well documented that changes in the chlorophyll a concentration (Chl) are accompanied by more or less systematic variations in the IOPs and AOPs, such as the spectral absorption coefficient a(A), the scattering coefficient b(X), the diffuse attenuation coefficient for downwelling irradiance K,(h), and ocean reflectance R(X) (e.g. Smith and Baker 1978; Gordon and Morel 1983; Morel 1988). The correlations between chlorophyll concentration and optical properties have formed the basis of statistically derived biooptical models that predict IOPs and AOPs given Chl, or vice versa (e.g. Morel 1988). Although such chlorophyll-based bio-optical models may often satisl’actorily predict average values as obtained from water samples taken at many locations and times, these models say nothing about the variability observed in different water samples, each of which has the same chlorophyll concentration. We here refer to this variability as random, which is used in a sense that the measured optical properties differ in a seemingly random manner from the mean values predicted by chlorophyll-based bio-optical models. As a result of this variability, any particular measured IOP or AOP can Microbial optical database 539 frequently differ by a factor of two or more from the value predicted by the bio-optical model. For example, data presented in Gordon and Morel (1983) show a three-fold variation in the scattering coefficient b(550) at any given chlorophyll concentration in case 1 waters. These random variations are much larger in case 2 waters, which include most coastal areas of the world’s oceans where components other than phytoplankton and their derivatives are optically significant (e.g. resuspended sediments, terrigenous particles, or dissolved organic matter). The data of Gordon and Morel show that b(550) may vary by one order of magnitude or more for a given Chl if case 2 waters are considered. A major motivation for the present work is provided by the hypothesis that the seemingly random variability in the chlorophyll-based bio-optical models can be explained in terms of the detailed composition of seawater. We think that much of this variability is due to the wide range of microbial particles (chlorophyll-bearing or not) and nonliving particles present in seawater. We thus expect that an improved understanding of the marine optical environment can be gained if we progress beyond the overly simplified parameterization of marine particulates in terms of chlorophyll concentration alone. The most rigorous approach to understanding the complicated marine optical environment begins with determining the optical properties of each of the various optically significant components of seawater. A definitive approach clearly should consider the optical properties of every species of aquatic components present in a water body, as well as the possible intraspecies variability in these properties. Such an approach is clearly unattainable because of the wide variety of particulate and dissolved species in natural water bodies. An alternative approach is to select a number of species or components that realistically represent the mix of suspended and dissolved matter that affects the marine optical environment. We are taking such an approach in the present work, which combines the single-particle optical database of various microbial particles with radiative transfer modeling. In contrast to correlational, chlorophyll-based bio-optical models derived from statistical analysis of (usually incomplete) field data, our approach leads to the development of a comprehensive mechanistic model of the marine optical environment. To utilize such a mechanistic approach requires knowledge of the optical cross sections of various kinds of aquatic components including microbial particles. These optical cross sections represent the amount of absorption and scattering that may be attributed to a unit concentration of a given component. In this work we specifically use the absorption and scattering cross sections that are attributed to a single particle representative of a given microbial species. These single-particle cross sections provide linkages between the bulk IOPs of a water body and the number concentrations of its particulate components. Importantly, if the optical cross sections and concentrations are known for a given species of particulate assemblage, it is possible to determine the contribution of this species to the bulk optical properties of the water body and hence to determine its effect on radiative transfer in that water body. Recent progress in two research areas has made it possible to undertake this study. First, there has been a considerable effort to study the optical properties of different types of microbial particles, from the smallest bacteria to various nanoplankton species (e.g. Morel et al. 1993; Stramski et al. 1995 and references therein). Second, a numerical radiative transfer model (Hydrolight 3.0) has been developed that is both accurate and computationally efficient enough to permit extensive studies of oceanic light fields as particle types, concentrations, and optical properties are systematically varied (Mobley 1994). This paper represents the first part of our work, wherein we describe the database of single-particle optical properties of several kinds of microbial particles covering the size range from -0.05 p,rn (viruses) to -10 km (nanoplanktonic chlorophytes). This database is a result of several recent studies that combined laboratory experiments and Mie scattering calculations, and includes spectral cross sections for absorption and scattering as well as scattering phase functions. Our companion paper (Mobley and Stramski 1997) is the second part wherein we describe the use of our database in radiative transfer modeling.