NEW TECHNIQUES

MATCHING THE TOTAL AND POLARIZED WATERLEAVING RADIANCE CONTRIBUTIONS TO THE MULTIANGLE AND MULTISPECTRAL REMOTE SENSING MEASUREMENT CAPABILITIES OF THE 2009 NASA/GLORY MISSION

Chowdhary, Jacek¹; Cairns, Brian¹; Travis, Larry¹; Mishchenko, Michael<sup>1

1</sup>NASA/GISS & Columbia University 2880 Broadway, New York, NY, 10025, United States

The polarization intensity of light scattered by particles exhibits features as a function of wavelength and of scattering angle that are distinctively different from the features for the total intensity of this light. The polarized and total intensity features exhibit also very different sensitivities to particle properties such as size, shape, and composition. Finally, these sensitivities vary themselves with particle properties. For example, the polarization of light scattered by low-refractive particles (relative to the surrounding medium) such as marine particulates shows less features with scattering angle, and less variation with shape and size, than the corresponding features for light scattered by high-refractive particles such as atmospheric aerosols. This suggests that for observations over the ocean, one can use the different variations of polarized and total intensity features with observation angle and wavelength to separate aerosol signatures from changes in the ocean color and retrieve both these atmosphere-ocean system properties simultaneously.

However, the success of such retrievals depends strongly on the ability to measure total and polarized intensity features accurately, and on the numerical tools to match such features for observations over the ocean. This sets the stage for his talk as follows. First, we introduce the multi-angle, multi-spectral measuring capabilities of the polarimeter onboard the NASA/Glory mission scheduled for launch in 2009. Secondly, we present the Case-1 water hydrosol model developed for this mission to compute realistic underwater light scattering contributions of total and polarized light to observations from space. Thirdly, we describe simulations of such contributions as a function of viewing angle, wavelength, and Chlorophyll a concentration, and illustrate the underlying potential for separating aerosol and ocean color retrievals. And fourthly, we discuss actual measurements and analyses of such contributions with an airborne version of the NASA/Glory polarimeter deployed at low (65 m) and high (4 km) altitudes.

EXPLOITATION OF MULTIANGULAR AND POLARIMETRIC REMOTELY SENSED DATA FROM SPACE FOR OCEAN COLOR RETRIEVAL IN OPEN OCEAN WATERS

Harmel, tristan¹; Chami, malik<sup>1

1</sup>LOV-UPMC LOV, B.P. 08 , Villefranche sur Mer, --, 06230, France

PARASOL (Polarization and Anisotropy of Reflectance for Atmospheric Sciences coupled with Observations from a Lidar) satellite instrument has been deployed in 2005 to provide complete observations of radiative properties of the atmosphere and ocean. The PARASOL sensor is currently the only one which is able to perform multiangular and polarized acquisitions combined with the spectral information. Our study focuses on the exploitation of the polarized and directional measurements carried out by PARASOL for remote sensing of the ocean color in open ocean waters. First, the variations in the polarized signal acquired in a visible spectral band (namely 490 nm) with water constituents are analyzed using radiative transfer modeling together with satellite data. The analysis reveals the property of invariance of the polarized satellite radiance at 490 nm with respect to phytoplankton concentration for the open ocean water case. Second, an original atmospheric correction algorithm is developed based on the latter results to improve the retrieval accuracy of oceanic parameters.. The new atmospheric correction algorithm consists in two steps. The first step focuses on the retrieval of atmospheric parameters using the influence of directional properties of the aerosols (including the non spherical aerosols) on the polarized and unpolarized radiance. the second step of the algorithm deals with the retrieval of the marine reflectance using a coupled atmosphere-ocean radiative transfer model that includes the polarization state of light. As a validation, the comparison of marine reflectances derived from one year PARASOL atmospherically corrected images with in situ measurements which are routinely collected by the BOUSSOLE buoy in the Mediterranean Sea, will be performed. Note that the proposed atmospheric correction algorithm might be extended to any sensors capable of measuring the directionality and polarization from space such as the Aerosol Polarimetry Sensor which will be shortly launched on the Glory (NASA) satellite.

POLARIZATION MEASUREMENTS IN COASTAL WATERS USING HYPERSPECTRAL MULTI-ANGULAR SENSOR

Tonizzo, Alberto¹; Zhou, Jing¹; Gilerson, Alex ¹; Iijima, Takako¹; Twardowski, Michael²; Gray, Deric³; Arnone, Robert³; Gross, Barry¹; Moshary, Fred¹; Ahmed, Sam<sup>1

1</sup>City College of the City University of New York 160 Convent Ave., New York, NY, 10031, United States; ²Department of Research, WET Labs, Inc., Narragansett, Rhode Island, 02882, United States; ³Naval Research Laboratory, Code 7333, Stennis Space Center, Mississippi, 39529, United States

Polarization characteristics of coastal waters are of great interest since they can be used in retrieval algorithms for the separation of organic and inorganic particulates, in improving underwater visibility, in understandingthe physics of light propagation of ocean lidar and in other active techniques and applications. To study these characteristics a new Stokes vector instrument has been developed by the Optical Remote Sensing Laboratory at CCNY. The instrumentusesthree hyperspectral Satlantic radiance sensors eachwith a polarizer positioned in front of it and with polarizationaxes aligned at 0, 90 and 45 deg. The sensors are mounted on ascanning systemwhich is rotated by a stepping motor, so the sensors canchange their angular positionin the range 0-180 deg in respect to the direction of the incoming sun light. Downwelling irradiance is also monitored by a fourth hyperspectral sensor positioned on the deck of the boat. Results of measurements of water polarization properties using this instrument during a recent cruise on R/V "Connecticut" in the coastal areas of New York Harbor- Sandy Hook, NJ region are presented for waters with chlorophyll concentrations 1-10 mg/m³, minerals concentrations1-2 mg/l, and CDOM absorption at 400 nm approximately 0.5 m^-1. Components of the Stokes vector and values of degree of polarization measured in the main scattering plane are compared with simulated ones using a Monte Carlo radiative transfer codefor theatmosphere-ocean system showing reasonable agreement.

OPTIMIZATION OF NEXT GENERATION OCEAN BIOLOGY REMOTE SENSORS USING A UNIFIED APPROACH TO FORWARD AND INVERSE MODELING

Stamnes, Knut¹; Li, Wei¹; Spurr, Robert²; Hamre, Boerge³; Stamnes, Jakob J.<sup>3

1</sup>Stevens Institute of Technology Castle Point on Hudson, Hoboken, NJ, 07030, United States; ²RT Solutions, 9 Channing Street, Cambridge, Massachusuetts, 02138, United States; ³Department of Physics and Technology, Bergen, Hordaland, 5008, Norway

For SeaWiFS, MERIS and MODIS, ocean color retrievals rely on simplified two-step algorithms based on an atmospheric correction followed by two- or three-channel regression to deliver chlorophyll concentrations. We have already shown that a one-step simultaneous retrieval of aerosol and marine properties based on linearized coupled atmosphere-ocean radiative transfer and optimized bio-optical modeling yields a considerable improvement in retrieval accuracy [Li et al., IJRS, in press, 2008]. Next generation ocean biology remote sensors should be optimized to retrieve the maximum amount of information from the radiances measured in space. Current ocean color sensors have channels in the visible and near infrared. Spatial resolutions are quite good, but spectral variability is limited, and they do not measure polarized light. Aerosol information obtained from space is greatly enhanced with polarized backscatter measurements and multi-angle observations of the same scene, which is important for ocean color, because most of the satellite signal comes from the atmosphere. We discuss how to apply systematic error analysis and information budgeting to determine the importance of both polarization and multi-angle measurements for ocean color retrieval, and thus find the best choice of state-vector aerosol and marine parameters to be included in a simultaneous least-squares inversion procedure. Error budgeting requires extensive and accurate simulations of backscatter measurements and associated sensitivities (weighting functions). For this purpose we use a suite of fully coupled atmosphere-ocean multiple scattering radiative transfer models with scalar (intensity only) and vector (with polarization) capabilities, and the linearization facility to deliver analytic weighting functions. Exploring the wavelength range from the ultraviolet (300 nm) through the visible and the infrared to 2,500 nm, we discuss how to optimize the wavelength channels, channel widths, and viewing angle capability for a dedicated multi-angle ocean color polarimeter.

LISST-BACK OBSERVATIONS OF NEAR-PI SCATTERING: STUDIES ON CONTRAST BETWEEN MONO-DISPERSE SPHERES AND RANDOM SHAPED GRAINS

Agrawal, Yogi¹; Boss, Emmanuel²; Mikkelsen, Ole<sup>3

1</sup>Sequoia Scientific, Inc. 2700 Richards Road, Suite 107, Bellevue, WA, 98005, United States; ²School of Marine Sciences, Orono, Maine, 04469, United States; ³2700 Richards Road, Suite 107, Bellevue, WA, 98005, United States

Backscattering near-pi is an important input to radiative transfer models used to study and interpret LIDAR data. A new instrument, the LISST-BACK, has been developed for the measurement of near-pi scattering of laser light. The instrument employs a 532 nm doubled-YAG laser, a CMOS array detector and optics similar to those used by Maffione and Honey (SPIE v.1750; 1992) and earlier by Kuga and Ishimaru (App. Opt. 1989). A key difference is that the present instrument incorporates a finite beam path of 20 cm so that studies of vertical variability of near-pi scattering may be observed. Further, the instrument measures beam attenuation, so that indirect estimates of c (Maffione and Honey, 1992) are not necessary. Here we describe the instrument and calibration tests performed with polystyrene beads, including measurements that reveal azimuthal variations consistent with the model of Pal and Carswell (App.Opt. 1985) and others. The instrument was deployed in a bottom boundary layer (BBL) experiment off the coast of Martha’s Vineyard, Massachusetts. Near-pi backscattering data from the BBL are compared with very sparse prior results. The observations reveal very little angular structure in the backscattering near pi. In contrast, Mie theory prediction using particle size distribution data from a nearby LISST-100X instrument suggests that a highly oscillatory structure in angular scattering should exist. It is suggested that this difference between observed near-pi scattering and prediction based on Mie theory is due to shape effects. Experiments with random shaped grains from an ISO standard dust sorted in a stratified settling column were conducted and their near-pi scattering is contrasted with equal size glass beads to illustrate shape effects. This study constitutes an effort toward replacing Mie theory to predict backscattering from measured size distributions, for LIDAR or remote sensing.

MARINE MAMMAL OBSERVATIONS USING MULTI-CHANNEL IMAGING SYSTEM

Schoonmaker, Jon¹; Contarino, Vincent M²; Gilbert, Gary¹; Podobna, Yuliya¹; Oakley, Daniel¹; Boucher, Cynthia<sup>1

1</sup>Advanced Coherent Technologies, LLC 4022 Liggett dr, San Diego, --, 92106, United States; ²Naval Air Systems Command, Patuxent River , MD, 20670, United States

The Advanced Coherent Technologies Mission Adaptable Narrowband Tunable Imaging Spectrometer (MANTIS) was used to detect and monitor marine mammals in the St Lawrence Seaway and in Maui Hawai’i. Both NADIR and grazing angle view angles were used. The system was configured with both narrowband spectral filters and polarization analyzers to reduce glint from the sea surface and multiple field of views in a foveal arrangement to explore both detection and classification in a single system. In addition optical models were combined with the diving profiles of various marine mammals to estimate the probability of detecting mammals as a function of water clarity and multi-channel system performance. Airborne and groundbased imagery of marine mammals are shown at various ranges and modeling results are presented.