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 wind ruffled 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 are presented for a realistic atmosphere-ocean system to show the effects of the volume scattering function, the dielectric interface, and waves on the induced errors.

During the past few years many methods have been proposed for estimating surface radiative fluxes (shortwave radiation, photosynthetically active radiation -- PAR) from satellite observations. We have developed algorithms for computing the shortwave radiative flux (shortwave irradiance) at the ocean surface from visible radiance observations and they have been found to be quite successful under most atmospheric and cloud conditions. For broken clouds, however, the simple plane parallel assumption for solving the radiative transfer equations may need to be corrected to account for cloud geometry. The estimation of PAR is simpler because the most commonly used satellite radiance measurements cover a similar region of the solar spectrum. We are in the process of producing global $ARDNSW and PAR as a contribution to the Sequoia 2000 project (to implement a distributed processing system designed for the needs of global change researchers). Results from our algorithms developed for Sequoia and preliminary global surface solar irradiance and PAR fields are presented and discussed.

A model to predict the monthly solar irradiance incident at the ocean surface has been developed at seven wavelengths across the visible spectrum (390, 440, 490, 540, 590, 640, and 690 nm). The model incorporates specific monthly climatological databases of aerosols, ozone, and percent cloud cover derived from satellite observation for the North Atlantic from May 1979. The variations in the spectral irradiance fields in the North Atlantic are shown to be highly spatially variable in small scales (< 100 km) in addition to being spectrally different. The irradiance distribution is dependent on the spectral characteristics of the transmittance parameters, and also on the small scale variability of the climatological data. The model indicates that the Rayleigh and aerosol transmittance has a pronounced affect on the spectral irradiance while the magnitude of the irradiance is controlled primarily by percent cloud cover and aerosol transmittance. The general trend in the North Atlantic indicated that the spectral irradiance intensity is similar to the solar spectrum (peaking at 490 - 540) and diminishing in the shorter and longer wavelengths.

Remote sensing reflectance is easier to interpret for the open ocean than for coastal regions since bottom reflectance and fluorescence from colored dissolved organic matter (CDOM) need not be considered. For estuarine or coastal waters, the reflectance is less easy to interpret because of the variable terrigenous CDOM, suspended sediments, and bottom reflectance, since these factors do not covary with the pigment concentration. To estimate the pigment concentration, the water-leaving radiance signal must be corrected for the effects of these non- covarying factors. A two-parameter model is presented to model remote sensing reflectance of the water-column, to which contributions due to CDOM fluorescence, water Raman scattering, and bottom reflectance have been added. The purpose of this research is to try to understand the separate contributions of the water-column, CDOM fluorescence, water Raman, and bottom reflectance for stations on the West Florida Shelf and Lake Tahoe. This model requires data with spectral resolution of 10 nm or better, consistent with that provided by AVIRIS and expected from HIRIS.

Under clear skies, strong fluctuations in the downwelling irradiance, E_d, prevail in shallow water as a result of the focusing and defocusing of sunlight by surface waves. Such temporal fluctuations were measured in the Black Sea, usually at a depth of 1 m, from a fixed platform located 600 m off the coastline. A method of thresholding analysis was applied to 109 time-series records of E_d (525 nm), each of which lasted 10 min. The frequency of occurrence of flashes of intense foci (intensity exceeding the time-averaged irradiance, E_d, by > 50%) decreased exponentially with increasing flash intensity. The frequency and intensity of flashes, hence the slope of the exponential relationship, all varied with wind-wave conditions and atmospheric lighting conditions. The best conditions for wave focusing were characterized by light winds of 2 to 5 m s^-1, solar elevation > 40 degree(s), and diffuseness of surface irradiance < 40%. Then, at a depth of 1 m, the flashes > 1.5 E_d occurred at rates as high as 6 Hz. The most intense flashes exceeded E_d 5-fold at rates of 10^-3 Hz. These results, which are consistent with our previous observations, substantially improve the database on still poorly documented wave focusing effects.

Subsurface heating rates from visible solar irradiance were computed for the North Atlantic for May 1979 during the period of the spring bloom. The visible spectrum between 400 and 700 nm accounts fora substantial fraction, about 43%, of the total solar irradiance at the sea surface, and comprises most of the solar irradiance that penetrates more than a meter into the sea. The mean monthly spectral diffuse attenuation coefficients and surface solar irradiance were used to compute heating rates at 390, 440, 490, 540, 590, 640, and 690 nm over depth increments of 0 - 5, 5 - 10, 10 - 15, 15 - 20, and 20 - 35 m. At low latitudes, in the North Atlantic, significant solar heating occurs at depth as a consequence of high solar irradiance and clear waters. In the northern latitudes the heating is confined near the surface at all wavelengths as a result of high turbidity. Significant spatial variation in the spectral heating rates is observed as a result of chlorophyll and aerosol patchiness.

A simulation model is described which generates simulated images of underwater objects as viewed through a wind-roughened ocean surface. The physical model includes representations for the two-dimensional wavy surface (gravity waves), beam spread at the surface due to small scale roughness (capillary waves), and beam spread and attenuation due to multiple scattering and absorption in the water. The sensor is modeled as a monostatic imaging system of arbitrary incidence angle, with emphasis on LIDAR systems. Results of the simulations are presented, illustrating the distortion of images in active seas, and the loss of resolution due to surface and volumetric scattering.

Optical imaging in an ocean environment is severely degraded in range and in resolution by distortions introduced by intervening media. Phase conjugation is a nonlinear optical technique that can correct media-induced distortions. This technique requires the high energy densities and coherence of laser radiation. Degenerate four-wave mixing in photorefractive crystals, such as barium titanate, can provide either externally pumped or self-pumped realizations of phase conjugation. The primary limitations in the application of the technique are the formation time of the corrected (i.e., phase conjugate) image and the amount of optical power required to form the image. The effectiveness of optical phase conjugate image correction was investigated for distortions due to (1) static variations in the refractive index, (2) temporal variations in the refractive index, and (3) an air-water interface. Fundamental issues related to the ability to correct for real time fluctuations and to the laser power requirements of the system were studied. We discuss the capabilities of phase conjugation to correct environmentally induced optical distortions in real time.

The exponentially varying atmospheric density near the water surface can bend the radiation path and potentially affect optical detection and tracking by varying the maximum inter-vision range (MIVR) by causing a positioning error and producing mirages. Using a marine boundary-layer model in conjunction with ray-tracing, quantitative analysis of these effects as a function of meteorological conditions can be achieved and predictions on the nature and magnitude of the induced phenomena can be made. This simple form of analysis produces effects of significant magnitude depending on the conditions. However, the literature reports very few instances of these effects and the few data published on the subject lack the necessary information to relate the phenomena to the prevailing weather conditions. An experiment was conducted over the Ottawa River in the fall of 1991 to verify the occurrence, persistence, and magnitude of these refraction-induced phenomena and initiate the validation of our modelling approach. Both shortened and extended MIVRs as well as mirages were observed and the measurements made were in good agreement with the model predictions. Sample images taken under sub- and super-refraction conditions are presented.

A lidar system, developed to map atmospheric extinction under the flight path of a P-3 aircraft, was tested during the Key 90 experiment. Using a modified Cassegrainian telescope, the return signals from both wide and narrow field of view are detected. Optical depth and extinction profiles are derived from these signals. The instrument design, data analysis, and measurements from Key 90 are discussed.

An aureole lidar technique for estimating extinction profiles has been developed at the Naval Research Laboratory (NRL), Washington, DC. This technique has been evaluated by making nearly simultaneous measurements of atmospheric extinction utilizing the NRL airborne Aureole Lidar Platform and the Naval Command, Control and Ocean Surveillance Center (NCCOSC), the RDT&E Division, Particle Measuring Systems, Inc. aerosol spectrometers. Profiles of measured aerosol-size distributions were used to calculate extinction coefficients and were compared with the aureole lidar estimations. Aureole lidar estimated extinction coefficient profiles are similar in structure to the PMS measured profiles and agreed well in extinction magnitude. When the near surface aircraft aerosol measurements are averaged over a longer time, excellent agreement exists between the aureole lidar and the PMS measured values.

An airborne optical receiver (AOR) was developed and tested to investigate the propagation and reception of optical communications uplinks from a submerged laser source to an overflying fleet aircraft. The AOR was flown in a P-3C Orion aircraft for an at-sea test off the southern California coast in August 1990. A green laser transmitter was suspended from the research platform FLIP at depths of 15 to 45 m. During six nights of operations, the AOR received the laser light at various test geometries and through clear and cloudy conditions. This represents the first optical uplink cloud experiment at visible wavelengths. Results show that optical pulses in clouds are significantly more forward-scattered than modeled. The results can be explained by Mie scattering theory. Measured cloud attenuation and pulse stretching agreed with an existing optical propagation model. Significant attenuation and signal spreading due to haze and fog was measured and compared with theory.

Measurements are reported of the infrared sunglint clutter channel caused by the direct solar reflection from the wave-perturbed sea surface at near-grazing angles of incidence. Apparent radiance has been measured over Monterey Bay as a function of azimuth and elevation angles relative to the sun direction using an AGA Thermovision 780 dual-band radiometric imaging system in the wavebands 2 to 5.6 (SW) and 8 to 12 micrometers (LW) with 7 degree(s) FOV. Time averaged profiles from multiframe averages show near-Gaussian angular distributions with half widths in the range 3 to 20 degrees (depending on solar angle) for look-down angles of 1 to 10 degrees below the horizon. The p- and s- polarized components of sea surface radiance have been obtained using an external wire-grid polarizing filter and compared with unpolarized measurements. The degree of polarization within the glint is shown to be horizontal and variable in the range 1% to 30%, depending on solar elevation, the higher degrees of polarization being found in the SW band. Significant vertical sea radiance polarization has been observed outside the solar glint in the 8 to 12 micrometers band, and is attributed to sea surface emission polarization.

The problem of retrieving spatial information of the sea surface heights from aerial images is considered. We proceed, for simplicity, by considering a one-dimensional model of the problem. With some simplifying assumptions, we derive some analytical and numerical results that relate the autocorrelation of the surface heights and those of the sunglint patterns. We assume that the surfaces are such that they constitute approximations to Gaussian random processes. We also assume that the surfaces are illuminated by a source (the sun) of a fixed angular extent and imaged through a lens that subtends a very small solid angle. With these assumptions, we calculate their images, as they would be formed by a signal clipping detector. In order to do this, we define a `glitter function,' which operates on the slope of the surfaces. To test our predictions we have conducted a Monte Carlo type simulation. Random surfaces with two different power spectra have been generated in a computer. We find that under favorable conditions, it is possible to invert the relation numerically and estimate the surface height autocorrelation from the sunglint data. We obtain the wave spectra from the surface height autocorrelation via a Fourier transform.

Optical techniques to measure the small-scale shape, i.e., the short wind waves of the ocean surface are theoretically reviewed. The well-known `shape from shading' and `shape from stereo' paradigms from computer vision are applied to a specular reflecting surface such as the ocean surface and used to study a variety of techniques with a common and elegant concept. The analysis shows that all techniques which have been used so far to take images of short wind waves such as Stilwell photography and various stereo techniques have significant deficiencies. Techniques based on light reflection (`shape from reflection') are basically only useful to derive wave slope statistics. A technique has been developed -- using an artificial light source to measure the 2-D probability density function of wave slope -- which is an extension of the successful sun glitter technique of Cox and Munk. Stereophotography is plagued by insufficient height resolution for small waves and, even more troublesome, by the problem that features seen in one of the images are not necessarily found in the other (correspondence problem) due to the specular nature of reflection at the water surface. Techniques based on light refraction (`shape from refraction') turn out to be most suitable to take wave slope images. They have been successfully used in the laboratory, but will be applied to the ocean in the near future.

The modulation of the spectral density of wind-waves 4 mm to 12 cm long by sinusoidal surface currents is measured using a scanning slope sensor. The results indicate that the magnitude of modulation decreases as wind speed or wavenumber increases, in qualitative agreement with the prediction of the relaxation mechanism. Quantitatively, the linear relaxation theory was found to under-predict the magnitude of modulation.

A thorough understanding of the hydrodynamics of short ocean wave is important for interpreting measurements made by active microwave remote sensing instruments. However, conventional methods for studying the structure of a water surface are not capable of resolving the fine scale structure of the surface, especially in the ultra-gravity and capillary wavelengths. Optical instruments have the potential for resolving the fine-scale structure of the ocean surface, however, methods for calibrating these instruments and verifying the accuracy of the measurements have not been developed. In this paper we describe a multi-faceted approach for verifying the accuracy and calibration of an imaging wave slope gauge (ISG). The first step is a thorough theoretical analysis of the geometrical optics and photometry. A detailed discussion on the relationship between surface slope and observed pixel intensity is presented. This discussion includes second order effects which may tend to bias the results. Secondly, calibration objects formed from thin transparent Perspex sheets with known slope and height profiles are retrieved. The results show that the measurements of the water surface shape are accurate enough to compute 2-D wave number spectra.

In this paper, we describe TRW's latest version of the scanning laser slope gauge (SLSG) which was used to characterize the sea surface in a recent ocean experiment. The SLSG, capable of measuring the spatial distribution and temporal evolution of the surface slopes of a patch of ocean, provides ground truth data which form a quantitative basis for the understanding of ocean wind-wave interactions and the development and validation of radar scattering models relevant to ocean remote sensing.

A detailed study of 2-D wave number spectra of short water surface waves is presented. Using a refraction-based optical technique either the along-wind or the cross-wind slope is visualized in image sectors of up to 30 X 40 cm². The resolution of the images is high enough (down to 1/3 mm) to resolve even the smallest capillary waves. The measurements were performed in the wind/wave facility of the IMST (University of Marseille, France) at 5 through 29 m fetch, the Delft wind wave flume (The Netherlands) from 6 to 100 m fetch, and the 4 m-diameter circular wind/wave facility of the Institute for Environmental Physics at the University of Heidelberg (Germany). A first preliminary analysis of the data is given. The angular dispersion of the waves is most sensitively influenced by the geometry of the facility, especially the width of the water channel. Therefore, it is hardly possible to extrapolate the measured angular dispersion of the waves to the ocean. The unidirectional and along-wind wave number spectra, however, show clear trends which allow for an extrapolation to the ocean. At high fetches and wind speeds, the spectral densities for the wave height are proportional to (kappa) ^-3.5 well into the capillary wave region until a sharp and almost wind speed independent cutoff occurs at (kappa) approximately equals 1100 m^-1 ((lambda) approximately equals 0.6 cm). The increase of the spectral densities with friction velocity depends both on wave number and fetch. While the spectral density for small gravity waves depends only weakly on the friction velocity, it increases strongly at higher wave numbers. Generally, this steepness is smaller at higher fetches.

We describe the design, implementation, and deployment of a laser slope gauge developed at the Woods Hole Oceanographic Institution for the purpose of studying the propagation characteristics of capillary ripples, and how currents and natural slicks on the ocean surface modify ripple spectra. The laser slope gauge constitutes a nondisruptive optical technique for determining the slope spectrum for a range of waves with wavelengths between 2 mm and 20 cm using both spatial and temporal information. Operation of the sensor and data acquisition system is discussed and a sample data record collected in the Gulf Stream off Cape Hatteras, North Carolina is interpreted and analyzed.

A description of a new scanning laser slope gauge is given and the preliminary results obtained from laboratory wave tank measurements are presented. The device relies on the measurements of two components of surface slope to compute spatial and temporal lags used to estimate the full three-dimensional spectrum. The device is capable of resolving frequencies in the range between zero and 63 Hertz, and wavelengths in the range between 0.63 and 20 centimeters. The technique makes use of a two-dimensional laser scanner which samples the perimeter of a 10 centimeter square (an unfilled aperture.) The laboratory results show mechanically generated waves propagating on both distilled water and a one micro-molar solution of Triton-X100^R in distilled water. Results indicate the device is well suited to measure the full three dimensional spectra of capillary-gravity waves and is capable of providing ground-truth measurements for the verification of remotely sensed ocean surface features.

In many remote sensing applications the shape of a water surface must be determined instantaneously, and under natural lighting conditions. Because of their nondestructive measurement capabilities, optical remote sensing systems that employ stereo image analysis techniques seem to be ideally suited for this application. Most stereo analysis techniques, however, assume surface reflectance properties which are incompatible with the reflectance properties of water, especially in the centimeter and smaller wave lengths. To analyze the fine scale structure of the ocean surface, a new specular surface stereo technique is presented that makes use of the unique optical properties of water. The method analyzes an image of the illumination source and multiple images of the water surface, and incorporates an image formation model that predicts the irradiance at a pixel for a given surface shape and illumination source. The surface shape is determined by solving the inverse problem of finding a surface elevation and gradient map that will result in a set of synthetic images that closely match the observed images. The performance of the specular surface stereo technique and a conventional stereo analysis technique were tested by processing simulated data. Comparison of the two techniques showed that the specular surface stereo method can potentially recover significantly more information than conventional stereo images analysis methods.

A new optical instrument has been designed for combined slope/height measurements of the small-scale structure of the ocean surface. The compact and rugged sensor head contains two light sources and a short-base CCD stereo camera setup mounted 4 - 6 m above the water surface and looking straight down onto the water surface. It takes stereo images of the specular reflexes on the water surface representing slope zero-crossings in a sector of about 30 X 40 cm². The height of the reflexes can be determined with a precision of about 2 mm. Experiments have been performed in the wind/wave flume of Delft Hydraulics, at the Scripps pier, and at the Noordwijk research platform in the North Sea. In these campaigns, a total of about half a million stereo images have been taken with continuous time series of up to 8 min at 30 frames/s. Some preliminary results are shown.

A method is provided for the decoupling of the upwelling light field from the bottom reflectance component. A principal components analysis of the remaining water volume backscatter component is performed. A method is suggested for a decoupling procedure for use in remote sensing over specific locales.

Coastal seawater was collected on incoming high tides at the Rosenstiel School of Marine & Atmospheric Science. 3.51 of water was transferred to a plexiglass bubble-tank and aerated for 30 sec at a flow rate of 4.0 ml/min per cm² of water surface area (158 cm²) using glass frits producing bubbles of 203 +/- 61 micrometers diameter. The surface pressure was then determined using calibrated spreading oils of known spreading pressure. After cleaning the seawater surface thoroughly, the water was re-aerated and allowed to stand for 1 to 3 hours in the absence and presence of longwave (365 nm) or shortwave (254 nm) ultraviolet (UV) light, both having an intensity of approximately 300 +/- 25 (mu) W/cm². In the absence of UV radiation, the surface pressure fell to approximately 64% of its initial value after 1 hour of standing and to approximately 63% of its initial value after 3 hours. Comparable results were obtained in the presence of longwave UV exposure. Under shortwave UV radiation, the decline in surface pressure was substantially accelerated; becoming 55% of the initial value after 1 hour and 35% of the initial value after 3 hours.

Imagery acquired by viewing through the air-sea interface is generally degraded by refraction at the interface, as well as by atmospheric and in-water scattering. A key problem in the restoration of such imagery is the removal of interfacial refractive effects. We have previously shown the feasibility of image restoration via model-based compensation for interfacial refraction and volume scattering distortions. Our techniques are contingent upon the accurate determination of sea topography, which we propose to accomplish via pointwise optical time- domain reflectometry (OTDR). Recently, consideration has been given to interfacial sensing methods such as glint imaging and stereophotogrammetry, which can be implemented with passive sensors but exhibit deficiencies which limit their utility in sensing sea topography. In this paper, we analyze salient errors inherent in the foregoing techniques, and briefly discuss errors incurred by sonic and microwave echolocation. We show that OTDR-based sensing is superior to stereophotogrammetry, whose accuracy, range, and field of view is resolution- limited. Additionally, we analyze and demonstrate the effect of topographic sensor errors upon feature localization in the reconstructed target imagery. Given a simulated wave height of 1.5 meters, 100 m sensor altitude, and 5 m target depth, we show that our image restoration model is stable, and remains useful, when the estimate of sea surface elevation is corrupted by a ranging error of up to 25.5 cm (peak-to-peak) at an error cross-section of 25 percent or less.

In a new application of stereophotography, the three components of the surface current can be obtained directly. The method does not require the deployment of an instrument in the water, or extrapolation of a subsurface measurement to the ocean surface. The method is applied to data collected in the North Sea and the results are shown to be in good agreement with the predictions of linear theory for surface waves. It is anticipated that the technique will have useful applications for measuring surface drift and currents induced by the orbital velocities of breaking waves, which play an important role in air-sea interaction at the ocean surface.

Optical properties of the atmosphere over the ocean are essentially different from the optical properties of the atmosphere over the land. For a cloudless atmosphere this difference is determined by the mechanisms of aerosol formation. For cloudy atmospheres this difference results from different mechanisms of convection. Convection over the sea is essentially weaker than over the land and the cumuli-form of clouds over the sea is essentially thinner. We must take into account the differences between optical properties for the atmosphere over the ocean and over the land for the purposes of remote sensing, LIDAR investigations, and in calculations of the radiative energy transfer between ocean and atmosphere. Taking this into account, it becomes apparent that we must correct all preceding atlases and maps that do not reflect this difference. It is also important for engineers who design optical instruments for use in the atmosphere over the oceans to consider the implications of this difference.