Virtually all calculations to date dealing with radiance calculations in an atmosphere-ocean system have been performed using a scalar theory approach where polarization effects have been neglected. This approach is always in error; however, neither the nature nor the magnitude of the errors induced has been studied. We have written a large scale Monte Carlo program to calculate the complete four component Stokes vector at any region in a fully inhomogenous atmosphere ocean system with inclusion of a stochastic interface. The program uses as input the Mueller matrices for both the aerosols in the atmosphere as well as the hydrosols in the ocean. The Mueller matrix for the stochastic interface is also accurately accounted for. The correlated sampling technique is used to compute radiance distributions for both the scalar and the Stokes vector formulations in a single computer run, thus allowing a direct comparison of the errors induced. Results will be presented for a realistic atmosphere-ocean system where Rayleigh scattering is assumed for both the atmosphere and ocean

A phenomenological model of the beam spread function is fitted to Duntley's measurements in artificial seawater in the turbidity range 0.60 less than or equal to s/α less than or equal to 0.83 which is constructed by first delineating the angle-dependence of the curves of the irradiance distribution for given radial distance and then obtaining the radial dependence of their scaling by normalizing to the total spherical-cap integral. The paraseters of the sodel are all easily identified graphically and therefore lend themselves readily to physical interpretation. The agreesent of the sodel with the data is excellent, and it is expected that it should extrapolate well not only to all distances and angles but also to values of the turbidity s/u well outside of the range of the data, including the high-turbidity halt 0.9 s/α 1.

Two simple algorithms are examined for estimating the volumetric generation rate of a bioluminescent light field in the upper ocean from measurements of the irradiance and scalar irradiance at two depths beneath the surface of seawater. Both algorithms require that in situ irradiance and scalar irradiance measurements be made and that the absorption coefficient be known. Radiative transfer calculations have been done to numerically test the sensitivity of the algorithms, and the one based on the principle of photon conservation is shown to work well when searching for the presence of a source and for estimating the source magnitude.

The radiance distribution in a three dimensional volume of ocean water is obtained using a numerical finite difference scheme applied to the radiative transfer equation. The solutions to the model are obtained for a sunny sky and a surface ship that creates a shadow in the water. The model simulates the conditions for which data was collected. Radiance, upwelling and downwelling irradiance, water reflectively, and diffuse attenuation coefficient from the model and the experiment are compared.

The Coastal Zone Color Scannez (ZCS) and associated atmospheric and in-water algorithms have allowed synoptic analyses of regional and large scale variability of bio-optical properties [phytoplankton pigments and diffuse auenuation coefficient K(490)}. Austin and Petzold (1981) developed a robust in-water K(490) algorithm which related the diffuse attenuation coefficient at one optical depth [1/K(490)] to the ratio of the water-leaving radiances at 443 and 550 nm. Their regression analysis included diffuse attenuation coefficients K(490) up to 0.40 nm, but excluded data from estuarine areas, and other Case II waters, where the optical properties are not predominantly determined by phytoplankton. In these areas, errors are induced in the retrieval of remote sensing K(490) by extremely low water-leaving radiance at 443 nm [L_w(443) as viewed at the sensor may only be 1 or 2 digital counts], and improved cury can be realized using algorithms based on wavelengths where L_w(λ) is larger. Using ocean optical profiles quired by the Visibility Laboratory, algorithms are developed to predict K(490) from ratios of water leaving radiances at 520 and 670, as well as 443 and 550 nm.

A model devised earlier1,2 relating the spectral dependence of K(.), the attenuation coefficient for natural diffuse light in the ocean, to the value of K()) at a reference wavelength, is re-examined using recently acquired data. The original model was based on data obtained from 1 967 to 1 979 by various investigators using a variety of scanning submersible spectroradiometers in tropical to mid-latitude waters. New data were obtained by the Visibility Laboratory in a series of cruises from 1985 to 1987 covering ocean latitudes from 24.4° to 77.4°. These data are believed to be of uniformly high quality and have the advantage over the previous data in that they permit the determination of K()) profiles to 200 meters from data obtained simultaneously at all 12 wavelengths and over a time span of approximately 6 minutes, vice data recorded serially at each wavelength and at discrete depths over a time span of as much as one to two hours with the spectroradiometers. An examination of the variation in the spectral K(A)'s with latitude was prompted by a concern that the phytoplankton species distribution at high latitudes might differ sufficiently from that at tropical to mid-latitudes to require a change in the spectral K(.\) model. A comparison was also performed between 83 sets of spectral K(A)'s derived from the new data and K(X)'s predicted by the model using measured K(490) as an input index. Small systematic differences were found which are not believed to be significant for most applications, hence, no change in the model is recommended on the basis of these findings.

The limits and nawre of the influence of water Raman scattering in clear ocean waters are the subject of continuing investigations. At shorter wavelengths (less than 500 nm) the in-water effects ofwater Raman scattering are confined to the near surface layers while the water leaving irradiance is significantly augmented by Raman emission. The water leaving Raman irradiance is apparently affected by the Fraunhofer lines of the solar spectrum.

The usual goal of aircraft and satellite remote sensing is to extract information which is directly related to ground targets in spite of atmospheric degradation which often complicates target identification and classification. Empirical algorithms which attempt to characterize targets by their spectral shape (slope, curvature, etc.) have been successful under special conditions, but fail when spectral variations in the solar or atmospheric parameters overwhelm those of the target reflectance. It is possible to derive an algorithm based on derivatives of the radiative transfer equation. This makes it possible to define the conditions under which a derivative algorithm will be insensitive to atmospheric effects and allows estimation of expected errors. This paper describes the development of the "derivative ratio algorithm," based on derivatives of a simple radiative transfer equation. The limiting conditions of the algorithm are derived and demonstrated using examples of reflectance spectra of turbid water and an ash leaf.

The design and calibration of a proposed in situ spectral absorption meter is evaluated using a laboratory prototype. The design includes a silver coated (second-surface) glass tube, a tungsten light source (stabilized by means of optical feedback), a monochromator, and a solid state detector. The device measures the absorption coefficient plus a portion of the volume scattering function. Theoretical analyses and laboratory experiments which explore the magnitude and variation of the errors due to scattering and internal reflections are described. Similar analyses are performed on the Cary 1 18 Spectrophotometer to allow cross calibration. Algorithms to yield the abscrption coefficient and the zenith-sun diffuse attenuation coefficient are presented and evaluated. Simultaneous measurement of the beam attenuation or backscattering coefficient allows use of algoriThms with much narrower error bands. The various methods of obtaining absorption and diffuse attenuation values are compared. Procedures for using reverse osmosis filtration to produce a clean water calibration standard are described. An absorption spectrum for pure water is obtained. Development of the absorption meter is proceeding along two lines: 1) a two-wavelength side-by-side LED is being fabricated to allow an in situ chlorophyll a absorption meter to be constructed, and 2) scientific projects using a shipboard or laboratory flow.-through pumping system are being planned.

Particulates in aquatic environments dominate variability in inherent and apparent optical properties. Dilute concentrations have made the quantitative determination of particulate absorption a relatively difficult problem. A quantitative filter technique (QFT) has been described previously1 ,2 which allows determination of the particulate absorption coefficient (a(A)) for samples concentrated on filters by correcting for the pathlength amplification effect (,8). DetaiI of the generality, accuracy and precision of the procedure have not been reported previously. A spectrophotometer equipped with an integrating sphere accessory was used to study the optical density of phytoplankton suspensions and colored polystyrene beads, OD(A), and the optical density of the same suspensions on a variety of common filter types, OD(A). The ratio of OD(A)/OD5(A) confirms that multiple scattering in filters leads to variable 6, a non-linear function of OD.(A). Studies of OD(A) when the filters are mounted in different positions in a single spectrophotometer, and for standard and integrating sphere spectophotometers indicate that the procedures presented are of general applicability. With proper care in baseline correction and sample preparation, spectra of a(A) can be calculated from measurements of OD\) for aquatic particles of diverse size and refractive index with an accuracy of better than i A comparison of filter types commonly used for aquatic research indicates that different algorithms are required for estimation of a(A) from OD(A) measurements using different filter types.

Light absorption in the ocean has been the least studied optical property because of the difficulties in making accurate measurements. With the previously used techniques, large differences have been reported for the specific absorption coefficient of phytoplankton (cultures and natural assemblages). It is difficult to determine if the diversity in these values are methodological or a function of actual variations in absorption. With the renewed interest and activity in optoacoustic spectroscopy (OAS), which accurately measures absorption, some of these discrepancies should be resolved. In this method, as molecules and particles absorb light from a modulated source, they thermally expand and contract, thereby generating acoustic waves, at the modulation frequency, which are detected by a hydrophone. Optoacoustic spectroscopy is ideally suited for measuring dissolved organic material and particle absorptions because of its high sensitivity (105m1) and the egligible effect of scattered light. In this paper the instrumental design for an optoacoustic spectrophotometer (OAS), which pecifically measures phytoplankton absorption (420-S5Onm), is described. The spectral absorption of dissolved organic material and a phytoplankton culture is presented. OAS holds promise in being able to measure absorption without use of either filtration or concentration techniques.

Particles from both natural populations and culture mixtures are categorized into microautotroph and detritus/microheterotroph components based on size and chlorophyll autofluorescence. Subcomponents ofmicroautotroph pigment groups are possible by the addition of phycoerythrin autofluorescence measurements. Subcomponents of microautotrophs as "live" and "dead" cells are determined by the addition of the metabolic activity stain fluorescein diacetate (FDA). The detritus/microheterotroph fraction includes green-fluorescing cells, abundant in the Gulf of California, detritus, and bacteria. The relative contributions of icroautotrophs, microheterotrophs and detritus mustbe discerned in order to fully interpret beam attenuation and changes in beam attenuation.

Accurate measurement of the optical absorption coefficient for both pure and natural waters has been complicated by the effects of scattering particulates. A new absorption meter which eliminates the systematic effects of scattering has been developed. This device, hereafter referred to as the Integrating Cavity Absorption Meter or ICAM, offers the advantages of being simple, sensitive, rugged, and capable of In situ measurements. The theoretical basis is sound and well oc1 Five versions of the ICAM have been constructed, calibrated, and tested. Effects of scattering have been tested experimentally and have been found to be negligible up to very high concentrations. Very precise results for aqueous suspensions are obtained when the absorption of pure water is assumed and is used for the calibration. However, using a less rigorous calibration procedure the absolute absorption of pure water can also be obtained. The results for the absorption spectrum of pure water throughout the visible compare favorably with accepted values.

Absorption spectra (400 to 700 nm) measured with a new instrument, the integrating cavity absorption meter, were compared to those measured with standard techniques for particles (collected on glass fiber filters) or solutions (in long pathlength cuvettes) . The integrating cavity absorption meter has two advantages over previously used methods: 1) low absorption signals can easily be measured due to the long effective pathlength created by the highly reflective walls of the cavity, and 2) scattering averages to zero because of the isotropic light source. For particulate absorption, the locations of absorption peaks were offset up to 5 nm in the integrating cavity spectra due to the use of interference filters for determining wavelength. The filter wheel has bands centered every 10 nm, while the spectrophotometer has a monochromator with 1 nm spectral resolution. Modifications of the integrating cavity, presently in progress, will include replacement of the filter wheel with a monochromator. Particulate absorption measured with the integrating cavity was similar to that measured with the glass fiber filters through a portion (500 to 700 nm) of the visible spectrum. However, from 400 to 500 nm, absorption measured in the integrating cavity was less than that measured on glass fiber filters. For example, at 400 nm, cavity values were only 45 to 83% of the glass fiber filter values. Possible causes of this difference are discussed. Absorption spectra for dissolved materials measured in the integrating cavity were similar to those measured in the spectrophotometer.

Measurements were taken from a mooring in order to address questions concerning variability of bio-optical and physical properties in the upper ocean. The measurements are relevant to questions involving: 1) the identification of bio-optical processes and their time scales; 2) relationships between bio-optical variability and physical forcing; and 3) the development and testing of coupled bio-optical and physical models. The concurrent measurements provide time series of beam attenuation coefficient, chlorophyll fluorescence, photosynthetically available radiation, and dissolved oxygen along with horizontal cutrents and temperature. Data were obtained during three consecutive mooring periods in the upper 160m from 7 to 8 depths in the Sargasso Sea (34N 70W). Sampling was done at 4 minute intervals during a nine month period in 1987. Here, data obtained from the first deployment (March 1 through May 10) are described. Some of the primary observations include: 1) the abrupt onset of springtime stratification and episodic changes in the beam attenuation coefficient and chlorophyll fluorescence; 2) advective water mass variations associated with a cold core ring and warm Gulf Stream outbreak waters in the vicinity of the mooring; and 3) diurnal variations in the near surface beam attenuation coefficient and chlorophyll fluorescence which are associated with daily cycles of biological primary productivity. The present in situ, high frequency, long-term observations provide an impetus for similar future observations relevant to studies of optical property variability, primary productivity, and particulate fluxes.

Concurrent time series measurements including: percent transmission of a collimated beam of 660 nm light (converted to beam attenuation coefficient, C), photosynthetically available radiation (PAR), and chlorophyll (chl) fluorescence were obtained at 7-8 depths within the upper 160m of the Sargasso Sea (34N 70W) from Mar-Oct 1987 using a moored array of instruments (see Dickey et al., this volume). A subset of these data, the spring deployment (the first of three) from March to mid May, has been analyzed with respect to the diel phase variations in the bio-optical properties. Among the features noted, the relationship between PAR and chl-fl changed from cM-fl lagging the PAR signal by ca 90 degrees, through 180 degrees of shift to leading by -90 degrees. The transition period corresponding to this change was marked by inconsistent behavior in thephase relationships between other bio-optical variables, but the changes were short-lived and returned to their previous offsets from the daily PAR. These changes are thought to be the result of a succession of species caused by a combination of warm outbreaks of Gulf Stream waters importing a foreign particle assemblage into the Sargasso Sea, contemporaneous with wind events causing deep mixing. This hypothesis is consistent with recent observations at sea and in laboratory studies.

In modeling the optical properties of and radiative transfer within ocean water, a common assumption is that the optical properties of the various dissolved and particulate constituents per unit concentration are constant. While this assumption is probably valid for non-living materials, it is frequently not the case with phytoplankton. Living cells are dynamic in their size, shape, and composition and respond readily (time scale of minutes to hours) to changes in the environment. Since each of these factors plays a role in determining the light scatter and absorption properties of the cell, the assignment of optical constants to phytoplankton cells, especially within Case I waters, may introduce significant errors in any theoretical treatment of radiative transfer and may result in erroneous interpretations of bulk optical measurements. Single-cell light scatter and beam attenuation of several species of marine phytoplankton, Thalassiosira pseudonana (clone 3H), an unidentified prasinophyte (clone OMEGA 48-23), Paviova sp. (clone NEP), Emiliania huxleyi (clone BT6), and an unidentified cryptomonad (clone 1D2), respond rapidly to increased light intensities. Flow cytometric determinations of cell refractive index n, measured at 488 nm, was found to decrease under high light conditions while cell size D, expressed as equivalent spherical diameter, increased. Beam attenuation c, measured at 660 nm, was generally found to increase when cultures were exposed to high light intensities. With a constant cell concentration, the observed change in n tends to decrease beam attenuation, while an increase in D tends to increase c. The magnitude and sign of dc/dt will depend upon the relative change in mean cell refractive index ii and the mean cell size i In response to bright light, i5f 3H and NEP was found to increase more rapidly than particulate organic carbon (POC). POC normalized to total cell volume increased within NEP and decreased within 3H', while at the same time, decreased within both cultures, suggesting that the cells swelled. As a result of light induced changes in i and ! optical properties of cells can change independently of biomass.

Optical properties of naturally derived bacterioplankton grown in unenriched seawater are described. The beam attenuation coefficient, absorption coefficient and size distribution of cells suspension of bacteria were measured in order to determine their optical efficiency factors. In addition, the bulk refractive index as well as the angular pattern of light scattering were obtained from Mie theory. The cellular scatering efficiency increases with decreasing light wavelength as A , the backscattering efficiency is almost spectrally neutral, and the absorption efficiency exhibits features associated with respiratory cytochromes. Except for backscattering, the efficiences are significantly lower than those for larger biological inicroparticles. We suggest that baceria are a xrtajor source of light scattering in oligotrophic waters, where their contribution to the scattering coefficient may far exceed 50%. This large contribution is caused by the fact that total geometric cross sectional area for the bacteria compensates for their lower scattering efficiency. The contribution by bacteria to particle absorption, although less certain, appears also to be important in oligotrophic waters, and it may even predominate a non-'phytoplankton component. The effects of changes in size and refractive index on optical efficiencies of bacteria are also discussed in terms of a recent hypothesis concerning the regulation of cellular water content.

Estimates of the in vivo specific absorption coefficients (m2 mg'; 400-750 nm, 2 nm intervals) for the major algal pigment groups (chlorophylls, carotenoids and phycobilins) are presented. "Unpackaged" absorption coefficients were initially obtained by measuring the absorption properties of pure pigment standards spectrophotometrically and "shifting" their absorption maxima to match in vivo positions. Two approaches for estimating the phytoplankton absorption coefficient (spectral reconstruction and spectral decomposition) are compared by linear regression analysis, incorporating concurrent measurements of particulate absorption and pigmentation performed in the Sargasso Sea. Results suggest that "pigment package" effects are minimal for natural assemblages of open-oceanic phytoplankton and that accessory pigments do not always co-vary with chlorophyll a over depth and time.

The modeling of oceanic remote sensing reflectance typically employs absorption and scattering parameters for the various constituents present in marine waters. Trans-spectral light sources such as fluorescence and Raman scattering are not generally parameterized in these models. Bioluminescence is not considered to be a significant contributor to water-leaving radiance measurements obtained mid-day, and has not been included in the models either. In this paper we present evidence of effects due to these three phenomena by comparing model results to remote sensing reflectances measured at several stations during the 1988 California Coastal Transition Zone (CTZ) Experiment. Differences between modeled and measured Rrs(A) values are discussed from the perspective of in-situ light source contributions.

The fluorescence emission (685 nm) from chlorophyll in phytoplankton makes a measurable contribution to the upwelling light in the sea. The emissions from other autofluorescing biological sources are not as well described. The photosynthetic accessory pigments (phycoerythnn and phycocyanin) comprise just one additional group of fluorescent substances. Many micro- and macroscopic heteroirophic marine organisms exhibit autofluorescent responses to incident radiation over all or part of their surfaces. The intensity may vary widely, and emission wavelengths span the whole visible spectrum. This paper will present microspectrofluorometric measurements of the fluorescence emission from a variety of marine subjects, ranging from single-celled members of the plankton community to both planktonic and attached invertebrates. As modeling of the underwater light field is refined to greater levels of detail there is a need for information on the precise wavelengths of fluorescence emission that might be introduced by biological sources, and on the abundance and distribution of those sources. The existing catalogue of observations of such sources is influenced by the constraints of epifluorescence microscopy and the widespread use of filter sets designed for optimal detection of chlorophyll. With appropriate measurement technique these limitations can be avoided and the catalogue expanded.

The predictive effects of Raman scattering on the inherent optical characteristics of aquatic water masses raises questions concerning our ability to estimate how light energy is budgeted. Using values for the optical cross-section of water and biogenic substances, we estimate the relative contribution of fluorescence from the microbiological component and Raman scattering to the underwater radiance field. We conclude that in oligotrophic conditions, the contribution by fluorescence or Raman scatter is equal , while in eutrophic waters fluorescence overwhelms Raman scattering.

Field tests were conducted during Aug/Sept 1989 of the second generation of a laser/fiber optic fluororneter attached to a microstructure proffling instrurnent (Rapid Sampling Vertical Profiler - RSVP). This instrumentation is designed to provide centirneter-scale biological measurements coincident with and at the same sampling frequency as those for temperature, conductivity, and horizontal velocity microstructure. The instrument used during the summer of 1989 employed a 200m dual fiber system, SMA optical fiber connectors used underwater, and an optrode (12mm diameter) containing two fibers, GRIN lenses, and filters. Shipboard laser excitation (488nrn or 514nm) was transmitted down a 2OOtm core excitation fiber, and the fluorescence emission spectrum returned via a 400pm core detection fiber to a shipboard multichannel array detector. Fluorescence emission spectra were acquired 30 times per second at an instrument drop rate of approximately 60 cm per second.

The Point Spread Function (PSF) is an importantproperty in predicting beam propagation and imaging system performance. An instrument to measure the in situ PSF of ocean water has been built and PSF profiles obtained. This instrument consists of two parts, a flashlamp with cosine emission characteristics, and an imaging solid state camera system. The camera system includes a thermoelectrically cooled CCD array with over 50dB of dynamic range. This allows the camera to measure the steeply peaked PSF over short (lOm) to long (80 m) ranges. Measurements of the PSF in three different locations are presented. One location was a coastal station off San Diego where the water column exhibited a well defmed shallow (approximately 30 meter) mixed layer with a particulate maximum (defined by a maximum in beam attenuation) at the bottom of this layer. During these measurements the PSF was highly variable with depth, as was to be expected due to the dependence of the PSF on particle concentration and size distribution. In the second example the water column was almost homogeneous (as evidenced in the beam attenuation profiles). Hence, the PSF showed very little dependence on depth. Measurements of the variation of the PSF with range are also presented. A simple relationship of the variation of the PSF with angle and optical path length is presented.

We have built a radiometer that measures ten integral moments of the radiance as functions of depth in natural waters. From these data it is possible to calculate nine spherical moments of the scattering function, providing this function varies slowly in horizontal planes (i.e. the water is fairly stratified). This technique inverts the equations of radiative transfer, which avoids some limitations of conventional instruments. The radiometer does not rely on small samples. Moreover, it measures coefficients that apply directly to radiative transfer, avoiding intercalibration problems from multiple instruments. The instrument is symmetrical about its vertical axis. Reflecting optics look through a cast acrylic tube 20 centimeters outside diameter and 1.6 meters long. Design depth is 150 meters. So far we have tested the radiometer in a pool. Soon we expect to make it seaworthy and take it on a voyage.

We have evaluated a number of spectral attenuation meter designs based on constraints related to power consumption, spectral bandwidth, sampling time, accuracy and stability . Our fmal instrument design employs a unique optical bridge deve1oped1r Sea Tech with ONR support, a tungsten light source and a holographic grating monochromatorThe instrument design is summarized as follows: White light from a 10-Watt tungsten lamp with a 1mm2 filament is collected by a condensing lens and then spatially filtered by a 1mm diameter pinhole which is placed at the entrance port of a monochromator. The monochromator has a 45°, 1200 lines/mm, holographic grating 37 mm in diameter with a 91 mm focal length. The grating is rotated about its vertical axis with a sine arm driven by a stepping motor, allowing wavelength to be selected from 400 to 800 nm. At the exit port of the monochromator we use a 1mm diameter pinhole which spectrally filters the output light, resulting in a spectral bandwidth of 9. 1 nm. This nearly monochromatic light is then measured by a unique reference detector with a 0.5mm diameter pinhole at its center, allowing light to be transmitted through the center of the detector. The transmitted light has a bandwidth of 4.5 nm. The monochromatic light is then collimated by a 50mm focal length achromatic lens and stopped down to a beam 1 cm in diameter. This light then enters the sample chamber. After passing through the sample the light is received by a 61mm focal length achromatic lens and is focused onto a signal detector with a diameter of 1.25mm. Digitized ratios ofreference detector to signal detector voltages allow transmission to be measured with an accuracy of 0.05% and a resolution of 0.01%. By monitoring temperature we were able to temperature compensate the instrument to within 0.05% transmission from 00 C to 25° C. Based on these results it is now possible to construct a spectral attenuation meter with the required sensitivity and accuracy to measure beam attenuation in water as clean as oligotrophic ocean waters.

As optical systems that make use of optical phase and optical phase rate become more common in underwater applications, so wiLl the need for a more complete understanding of the optical refractive variability of this medium. This paper provides an overview of the nature of instruments that are availabLe for this investigation and describes the preparations that are being made at the Naval Underwater Systems Center for the measurement of this physical property of the ocean.

Range-gated underwater imaging has long been a popular topic for discussion and analysis; however, current laboratory data from such a system in operation is not readily available. In the summer of 1989, the authors configured a range-gated, laser-illuminatedimaging system utilizing "off-the-shelf" equipmenttodemonstrate the feasibility ofusing this technique. Using an intensified, gated, solid state video camera, a pulsed-frequency doubled Nd:YAG laser, a time delay generator, and an optical delay to compensate for system delays, a range-gated testhed was configured with the capability to record image slices through the water volume movable in 1 ns increments. A time delay generator and an optical delay to compensate for fixed system delays were configured to gate the camera to take image slices through the water, moving in 1 ns steps. The tests consisted of placing the target at various depths to 40 feet as the turbidity of the water was varied and measured. The images were collected on video tape and later digitized and analyzed. Details of the experimental configuration and test results are presented.

In pulsed blue-green laser radar applications, the signal return amplitude can have a dynamic range of 9 orders of magnitude in 200 ns. Signal compression is required to match this range to the input capabilities of the digitizing or processing devices, typically 2 or 3 orders of magnitude. The compression of the dynamic range of the signal by photomultiplier space charge control (P5CC) results in an adjustable compression range, variable gain, bandwidth in excess of 100 MHz, and extended input dynamic range, with no loss of photomultiplier sensitivity (i.e. photon counting operations are still achievable).

The high frequency of optical radiation permits high data rate communications in the megahertz (MHz) regime. To the best of our knowledge, no unguided, underwater high data rate communications system has been reported. In this research, we investigate the feasibility of a short range, high data rate underwater laser communications system. The present study considers two independent systems that are separated by several attenuation lengths with regard to signal acquisition, communications channel lock, and information transfer. System self-alignment of the transmitter onto the receiver is demonstrated in a small laboratory water tank. Successful transmission of light modulated at MHz rates is demonstrated in a pool over an extended water path.

Computations have been performed of the incident spectral irradiance at the sea surface using LOWTRAN-7 as the basis to describe the incident scalar and vector irradiance in terms of the true solar zenith angle and the nominal visibility in the atmosphere. These computations have been used to describe the contributions to the incident irradiance from the direct and the sky components of the total irradiance and the average cosine of the sky component as a measure of the radiance distribution of the sky for varying atmospheric conditions. Comparisons of the computations from LOWTRAN-7 have been made with the results from other models, and with data obtained from field measurements, and excellent agreement has been obtained for the daily profiles of the vector and scalar irradiance at the surface. These computations have been used to provide a description of the irradiance at the sea surface for use in the analysis of remotely sensed data based on information on the radiative transfer through the atmosphere above the sea surface.

Mean and rrns fluctuations of downwelling solar irradiance below a rough ocean surface have been modelled using the small-angle approximation for the in-water radiance distribution. For comparison, a ray-trace approach has also been used to calculate the fluctuation statistics in the limit of no scattering. The mean irradiance decays rapidly near the surface, and at a sufficiently large depth (determined by the water optical properties) is within a few percent of the irradiance for a flat ocean surface. Although rims fluctuations decay with depth as well, their magnitude relative to the mean irradiance reaches a nonzero asymptotic value which depends on the sun position, optical properties, and surface roughness.

The effect of one-dimensional sinusoidal and trochoidal waves on the imaging of a submerged object is compared, Images were computed for crest-to-trough wave heights of 1/4 to 1 times the maximum possible value of 1/7, where 1 is the wavelength, and for wavelengths varying from 0.3 to 10 times the diameter of the object d for the passage of one period of the wave through the imaging frame for looking straight down. It is found that the slope discontinuity or near-discontinuity at the crest of the trochoidal wave results in a greater tendency toward fission of the image than is the case for the sinusoidal wave.

A simulation nodel is described which generates realizations of the LIDAR return from a wind roughened ocean, including both surface specular and subsurface volumetric returns. The physical model includes representations for the two dimensional wavy surface (gravity waves), beam spread at the interface due to small scale roughness (capillary waves), and beam spread and attenuation due to multiple scattering and absorption in the water. The sensor model allows for arbitrary incidence angles, transmitter divergences, and receiver fields of view. From ensembles of realizations, the statistical characteristics of the surface wave induced fluctuations are determined, such as profiles of variance versus depth, and spatial/temporal correlations of the returns. Model results are compared with experimental data on specular return statistics and downwelling irradiance fluctuations. Predictions are presented for the round trip LIDAR fluctuations induced by surface waves.

We have constructed a model which generates coherent time sequences of combined simulated whitecap and glitter radiance images. The algorithm uses a wiid wave power spectrum and generates an initial ocean wave surface via a Monte Carlo method. The linear wave dispersion relation is used to produce time sequences of ocean surfaces from the initial surface. Each of these surfaces is composed of facets that are defined by the numerical procedure. By treating each of the 'facets as a roughened surface, glitter radiance fields are generated. The reflectivity of each of the facets is a functional of the source and viewing angles, with roughness as an additional parameter. The wave height acceleration fields are also calculated from each of the ocean surfaces. A downwardacceleration threshold is used to determine whether a particular facet contains whitecaps and to determine the mean reflectivity. Thus whitecap radiance fields coherent with the glitter field are generated. The two radiance fields are then combined to generate simulated whitecap/glitter radiance fields which are compared to field data and results of these comparisons are presented.

During the seasonal transition from summer to winter conditions a profound transformation occurs in a sea ice cover. As air temperatures drop, the ice cools causing a reduction in the brine volume, melt ponds freeze, new ice forms in areas of open water, and the surface becomes snow-covered. There is a corresponding evolution in the optical properties ofthe ice cover with albedos increasing and transmittances decreasing. As part of the drift phase of the Coordinated Eastern Arctic Experiment (CEAREX), spectral albedos and reflectances in the visible and near-infrared (400-1100 rim) were measured during fall freeze-up. Observed albedos are presented for first-year ice, multiyear ice, and new-ice cases. In general, albedos increased as freeze-up progressed, with the increase being most pronounced at shorter wavelengths. There was a sharp increase in albedo associated with the surface becoming snow-covered. The greatest temporal changes occurred in a freezing lead where albedos increased from 0.1 for open water to 0.9 for snow-covered young ice in only a few days. The evolution of the transmitted radiation field under the ice was estimated using a simple two-stream radiative transfer model in conjunction with observations of ice morphology and thickness. Light transmission decreased dramatically due to ice cooling, snowfall, and declining incident solar irradiances.

A radiative transfer model designed for sea ice with vertical layering is presented which removes several of the computational problems associated with translucent media whose optical properties are strong functions of wavelength. Accurate results are achieved for both direct and diffuse incident irradiance for single scattering co-albedos ranging from 1.0 to l08. Calculated snow and sea ice albedos and extinction coefficients are consistent with both observations and previous models, and resolve certain apparent inconsistencies in previous observational results. Algae absorption has a significant effect on transmitted radiation but seems to be quite variable. Transmitted irradiances vary by more than a factor of 3 at certain visible wavelengths.

The physical properties of ocean water, in the top few ten meters, are of great interest in the scientific, engineering, and general oceanographic communities. Subsurface profiles of temperature, salinity, and sound speed measured by laser radar in real time on a synoptic basis over a wide area from an airborne platform would provide valuable information complementary to the data that is now readily available. The laser-radar technique specifically applicable to ocean sensing uses spectroscopic analysis of the inelastic backscattered optical signal. Two methods have received considerable attention for remote sensing and both have been demonstrated in field experiments. These are spontaneous Raman1 and spontaneous Brillouin2 scattering. A discussion of these two processes and a comparison of their properties that are useful for remote sensing was presented3 at SPIE Ocean Optics IX. This paper compares ocean remote sensing using stimulated Brillouin scattering (SBS) and stimulated Raman scattering (SRS) processes with better known spontaneous methods. The results of laboratory measurements of temperature using SBS and some preliminary results of SRS are presented with extensions to performance estimates of potential field systems.

Two Lambda Physik EMG2O3MSC excimer lasers were modified to pump a single dye oscillator cell providing a 500 Hz source for a Tuneable Airborne LIDAR (LIght Detection And Ranging) Transmitter. This modification to the standard Lambda Physik dye oscillator/amplifier design produced a significant weight reduction along with simplicity in alignment, both key aspects in an airplane environment. Previous modification included conversion to 400 Hz power, which increased beam divergence, but also increased the pulse energy. The Pulsewidth remained the same and the polarization had a preferred horizontal direction. We used Coumarin 102 dye as the lasing medium, providing tuneability from 470-510 nm with energy efficiences of 20% at the peak wavelength of 480 nm, which corresponds to a photon conversion efficiency of 31%.

Typical airborne Light Detection and Ranging (LIDAR) used for oceanographic measurements collect data at rates in the range of 1 Megabyte per second. This paper presents a method for organizing and displaying this great bulk of data to allow screening for areas of interest. In addition, this method can be transferred to dedicated hardware, to provide a 100% real time data display system at a reasonable cost. A typical airborne LIDAR system contains a scanning transceiver, and digitizes returned waveforms as the aircraft ffies some search pattern. Thus the data is inherently four dimensional (intensity and three spatial dimensions). This method reliesupon collapsing the four dimensional data into three dimensions; color and X,Y screen coordinates. This is doneby translating depth into an RGB color mix, and return intensity into RGB intensity. Thus color represents depth, and brightness represents signal strength. This data is then displayed on a high resolution color display. In order for this to be successful, some preprocessing is necessary to normalize the waveforms, so that changes in the displayreflect changes in the water column. In addition, some digital filtering is beneficial to increase the signal to noise ratio. This system is currentlyimplemented in Greenhills 'C' under CLIX (UnixVport toClipper) running on a Zaiaz 933 compute engine, with RASIL graphics software package on a Zaiaz FB 640 Graphics frame buffer. This hardware provides 5 MIPS average execution rate and 16.7 million colors on a 768 x 576 pixel display.

An algorithm is described for estimating wave heights and correcting measured local depths to the mean water level for an airborne laser bathymetry system utilizing collinear infrared and green beams. The wave heights are referenced to a mean surface derived from laser slant ranges plus vertical accelerometer data to isolate long period swell from aircraft motion. Performance of the technique under nominal operating conditions is characterized through detailed error analyses. It is seen that filters as long as several hundred seconds can be used to determine mean water level - as might be needed, for example, in the presence of swell crests parallel to the flight line. Use of the accelerometer also permits a sufficiently accurate mean water surface to be maintained for about 6 seconds without update by local surface slant range detections. This allows depths to be calculated for pulses whose surface returns are not detected. The use of slant ranges to calibrate system alignment parameters is demonstrated. Comparisons are made with an alternative approach used in the Australian LADS system.

Estimators of in situ bottom reflectivity are compared to laboratory measured spectra for several sites. Statistics are compiled on the relative fits of the estimators spectral curves to the laboratory data. A best estimator is selected based on these statistics. It is conjectured that the best in situ estimator remains the best estimate of the bottom reflectivity even in more complex ocean bottom areas where laboratory data may no longer provide accurate values for in situ sediment reflectivity.

Because of excessive suspended matter along coastal water of the China sea, it is very important to estimate the maximum penetration depth (MPD) of lidar for developing lidar bathymetry technique in Chma.This paper gives the relation between maximum penetration depth and Secchi disc depth from the radiative transfer theory and contrast transmission theory in water, and the distribution of MPD along the East China Sea.

It has been suggested that bathymetry measurements from multispectral digital imagery could be improved by basing the selection of bands to be used in the calculation on the water depth. Only the deepest penetrating bands should be used in the deep-water regions; only the longer wavelength, less penetrating bands should be used in the shallow-water regions. We have used Thematic Mapper bands one through four to test this new approach in an image of Key West, Florida. The results show no improvement over the standard algorithms. The method of band selection is described and bathymetric results are presented, including a comparison with the usual method of using all water-penetrating bands at once.

In June 1988, the Naval Ocean Research and Development Activity (NORDA) collected some "in-water" data using its Towed Underwater Pumping System (TUPS) in the near-shore waters off St. Andrews State Park, Shell Island, Florida. These in situ data include latitude; longitude; depth in meters; narrow-band upwelling at 465 nm, 507 nm, and 532 nm; broad-band downwelling collected at the surface; temperature; salinity; atid transmissivity. In this paper, we investigate the relationship between depth and the normalized upwelling irradiance (upwelling divided by downwelling) in the three bands. Algorithms used to calculate water depth from remotely sensed airborne and satellite multispectral data are applied to the TUPS data and results compared. The TEJPS data have the advantage over most aircraft- and satellite-collected data because they were collected over an essentially uniform bottom type (smooth sandy bottom with steady slope) and have no atmospheric contamination. A new algorithm for depth calculation is proposed.

A Monte Carlo method was employed to calculate scalar irradiance at 1 meter depth intervals, as a function of solar zenith angle and the ratio of scattering to absorption coefficients. Calculations were made for two angular scattering functions. Analysis yielded empirical equations by which scalar irradiance can be calculated at any depth with an error almost always less than 10%.

Using Monte Carlo simulations of radiative transfer in the ocean-atmosphere system, a scheme is developed in which the inherent optical properties of a homogeneous ocean can be estimated from irracliance measurements made near the surface. If just irradiance measurements as a function of the solar zenith angle are available, then the backscattering coefficient and the volume scattering function at scattering angles greater than about 600 can be estimated. If in addition the beam attenuation coefficient is measured, then the absorption and scattering coefficients can be estimated as well.

Cavity-enhanced photothernial spectroscopy has been applied to perform ultrasensitive absorption measurements of thin fims and liquid surfaces. Absorption coefficients of less than one part in 10 have been detected. Both spectral and time-resolved photothermal measurements have been carried out and the spatial resolution of the method has been investigated using a Ronchi ruling as the sample. The measurements are in qualitative agreement with our thermal modelling calculations.

Measurements of the intensity of the light scattered by 1 1 im polystyrene spheres suspended in water have been made as a function of angle from -0.05° to 0.05°. Data have also been obtained for scattering at O as a function of particle concentration. As the particle concentration Increases the sample transmission decreases monotonically but the scattered light intensity at O Increases monotonically as expected. This new measurement technique Is based on coherent two beam coupling In a photorefractive BaTIO3 crystal.

Eddies along the Polar Front/Marginal Ice Zone (MIZ) in Fram Strait are thought to make important contributions to nutrient flux and stimulation of primary productivity. During the Coordinated Eastern Arctic Regional Experiment (CEAREX) helicopter-based measurements of upwelling radiance were made in four visible spectral bands and in the thermal IR across mesoscale features associated with the MIZ. These structures were mapped by flying a grid pattern over the ocean surface to define eddy boundaries. Subsequently, the area was also sampled vertically with CTD and spectral radiometer profilers. Data obtained from a single structure were integrated to construct a three dimensional picture of physical and optical properties. Volume modeling of temperature, salinity, and density fields obtained from CTD survey define the subsurface eddy structure and are in good agreement with infrared derived characteristics. Maximum temperature in the core was found to be four degrees higher than the surrounding water. Volume modeling further indicates that a subsurface layer of Arctic Intermediate Water is intrinsically associated with the surface expression of the eddy. The ratio of upwelling radiances, L(44l)/L(565), was found to be correlated to surface chlorophyll, particulate absorption coefficient, and in water determinations of L using the optical profiling system. The remote sensing reflectance ratio along with the IR sea surface temperature were found to be useful to detect the surface expression of the eddy and to indicate near surface biological and physical processes.

This paper investigates the wavelength dependence of the polarization characteristics of light scattered from laboratory cultures of marine Clzlorella. Scattering measurements were obtained using a scanning polarization-modulation nephelometer at wavelengths of 457 and 514 nm. The experimental data are corrected for non-spherical contributions and the resulting curves compared to Mie calculations of coated spheres with a Gaussian size distribution. Although the absorption of Chiorella has been reported to be strongly wavelength-dependent in the blue to green region of the spectrum, the scattering behavior changes very little. To verify the sensitivity of the scattering technique to changes in the imaginary refractive index, measurements were performed on absorbing and non-absorbing suspensions of wellcharacterized, coated copolymer particles. In all cases, the angle-dependent measurements and calculations were compared for four elements of the 16 element Mueller scattering matrix at two wavelengths. In the past, comparison of scattering models and measurements were generally performed for only the total intensity (one element of the scattering matrix). The use of four elements provides a much more stringent test of scattering calculations than those based on a single element. Using this method we are able to infer information about the internal structure and refractive indices of microscopic single cell organisms in vivo.

The S11 and S14 scattering matrix elements were measured for light scattered from single dinoflagellates and single irregularly-shaped, alumina particles suspended in a transparent gel. The S14 matrix element indicates the degree of circularly-polarized light induced on incident unpolarized light. The S4 signal from the dinoflagellates was found to be significantly larger than that measured for the alumina particles. The nucleus of the dinofiagellate investigated, Prorocentrum micans, contains about 50 structurally complex, chromosomes with a helical structure. This work provides evidence in support of the hypothesis that the observed S14 signal produced by the light scattering from P. micans is due to the helical nature of their chromosomes.

An in situ fluorometer has been designed to determine the concentration of a bacterial-sized photoautotroph. The concentration may then be used to estimate their effect upon optical properties in the upper ocean. Their fluorescence signal at 575nm enables assessment of their natural abundance fluctuations. The fluorometer is designed to be used in profiling and in moored modes. Such an instrument can provide valuable information necessary for improving our understanding of the dynamics of bacterial-sized microorganisms within the particulate matter in the upper ocean.

This report provides a summary of the second in-water test of the LLNL/NAVSEA Underwater Laser Imaging System (UWLIS). The UWLIS is a laser-based, synchronous-scanning, underwater imaging device that is designed to operate at greater ranges than is possible with conventional underwater TV cameras. It differs from earlier prototype synchronousscanning systems in that it is capable of high scan rates that allow the generation of real-time, RS-170 video images. The UWLIS is being developed for eventual use on Remotely Operated Vehicles (ROVs) during deep-ocean Naval salvage missions. The floodlight-illuminated television cameras presently used on NAVSEA vessels can produce images at ranges up to about 2 attenuation lengths (AL) (2). Beyond that point, common volume backscatter from particulates in the intervening seawater between the imager and the target overwhelms the return signal, and the image is lost. The special optical geometry of the synchronous-scanning imager is designed to minimize common-volume effects. Previous theoretical studies (1) indicate that a system of this type should be capable of operation at distances as great as 6 to 7 attenuation lengths. An improvement of this magnitude would greatly increase the efficiency of salvage operations, thus decreasing their cost.

A method of discriminating radiance due to bright bottom reflectance from water volume reflectance has been developed for evaluating Coastal Zone Color Scanner products of chlorophyll concentrations and diffuse attenuation coefficients. A simple three-band thresholding algorithm is applied to CZCS water leaving radiance at 443, 520 and 550 nanometers. Thirty CZCS images were processed to characterize normalized water leaving radiances in three-channels. Statistics were obtained from reflectance signatures in areas of known bottom reflectance and areas of high volume scattering (significant sediment concentrations and low chlorophyll concentrations). High bottom reflectance areas located in clear shallow waters (optically shallow) of Bahamian and Florida Keys were found to have higher water leaving radiance in all three channels than areas of exclusively high volume scattering (optically deep), e.g. Mississippi Delta, Chesapeake Bay, and Sargasso Sea. Radiances exceeding thresholds in any of the three spectral channels are assumed to be corrupted by bottom reflectance and derived bio-optical products assumed invalid. Comparisons of bottom corrupted pixels are in close agreement with charted bathymetry and in-situ ship radiance data.

The influence of small waves (<2 to 10 cm in height) in in-ground tanks on the vertical attenuation coefficient for downwelling irradiance (Kd) calculated from either the mean irradiances over a 5 minute run or the average of point-to-point measurements made at 10 and 20 Hz is investigated. Waves of mean frequencies of about 1 Hz were introduced in the tanks that produced corresponding frequencies in the power spectrum of measured irradiances. Coherence was greater than 0.7 between spectra of irradiance and waves when small waves were introduced, but was less than 0.3 under ambient conditions. Both increases and decreases in Kd values were observed when comparing Kd calculated from the point-to-point values and that obtained using the mean irradiances. When large waves (5-10 cm) were introduced, the methods deviated by up to a factor of two, presumably due to the large fluctuations in the point-to-point Kd values relative to the fluctuations in irradiance averaged over the 5 minute run. There were changes of 30% in Kd, calculated from mean irradiances, as wave height was increased.

Measurements were made of the spread of a gaussian argon ion (489 nm) laser beam as it passed through up to 20 cm of sea ice grown in a water tank at air temperatures of -15, -25, and -35°C. These growth temperatures were selected to cover a range of ±10°C about the sea ice eutectic temperature of T_eut = -21.2°C. The Beam Spread Function (BSF) and the transform related Modulation Transfer Function (MTF) were obtained for increasing thicknesses of ice grown at these temperatures. In general the BSF could be represented as the sum of a gaussian distribution of narrow width and large amplitude due to multiple narrow angle forward scattering and a smaller wider spread associated with a growing diffuse light field from multiple wider angle scattered light. For ice grown at all temperatures rms spread was found to follow a 3/2 power law with ice thickness. The rms spread was found to be less in the ice grown at the colder temperatures due to increased attenuation with the smaller ice crystal size.

For some applications the results of atmospheric models can be used as inputs for the computation of underwater irradiance provided that the effect of transmittance through the air-water interface can be accurately estimated. In this work the computations of refraction and transmission through the air-water interface are reviewed, the effect of a variety of atmospheric types investigated, and a simple table provided for estimation of air-water transmittance. The propagation of spectral irradiance as well as the propagation of photosynthetically active radiation is considered.