Lisa H. Huddleston

Synthetic Image Generation of Shallow Waters Using a Parallelized Hyperspectral Monte Carlo & Analytical Radiative Transfer Model

Modeled hyperspectral reflectance signatures just above the water surface are obtained from radiative transfer models to create synthetic images of the water surface. Images are displayed as 24 bit RGB images of the water surface using selected channel. Comparisons are made in this paper between a hyperspectral Monte Carlo and a hyperspectral layered analytical model of radiative transport applicable to shallow water types. Images at the selected wavelengths or channels centered at 490, 530 and 680 nm suggest the two models provide the same results when displayed as RGB images. The most sensitive parameters for generating realistic images are water depth and bottom reflectance in clean natural, optically shallow waters. The images clearly demonstrate the need importance of detailed and accurate water depths.

Developing and testing a pushbroom camera motion control system: using a lidar-based streak tube camera for studying the influence of water waves on underwater light structure detection

A pushbroom sensor motion control system was developed for use in conjunction with a pulsed laser fan beam, streak tube camera, and a high speed low light level camera . The LIDAR and camera control system was tested to study the influence of water waves upon active-passive remote sensing systems and associated models that require pushbroom sensor motion. A pulsed laser fan beam signal at 532 nm was recorded using a streak tube camera and a (high speed, low light level, high quantum efficiency) digital CCD camera. Tests were conducted in 3 different water tanks, including 2 tanks with water waves (the longest wave tank or channel is 60 m long). Capillary waves, ~1 cm wavelength) were generated using an acoustic wave source generator. Streak tube camera and CCD images were collected in conjunction with a 532 nm pico-second short pulse laser. Images collected demonstrate the pulse stretching around submerged water targets as well as the ability to discriminate water depth of submerged targets in shallow water types. In turbid water, the pulsed layer backscatter structure showed a nearly random return as a function of depth if the signal was attenuated before reaching the bottom of the water column. The data collected indicated the motion control testing system can accommodate a variety of cameras and instruments in the lab and in the outdoor water wave channel. Data from these camera systems are being used to help validate analytical and Monte Carlo models of the water surface structure, and the underwater light field structure (pulse stretching) as well as to validate other LIDAR applications used in bathymetric and hydrographic surveys of coastal waters and marine inlets for physical and biological (submerged vegetation) surveys.

Iterative algorithm for layered optical remote sensing reflectance modeling of natural waters with depth-dependent aquatic constituent concentrations

This paper describes the radiative transfer of the sun’s electromagnetic energy utilizing a solution to the two-flow irradiance equations that generates fast and accurate estimates of light distributions in any layered media, such as water with depth dependent concentrations of water column constituents. The layered model is designed to generate synthetic water surface reflectance signatures and associated synthetic images, in the presence of depth dependent water constituents, various bottom types, and variable water depths. The layered model accounts for specular (collimated) irradiance below the water’s surface and utilizes boundary conditions that allow the absorption, backscatter, beam attenuation, and conversion (from specular irradiance to diffuse irradiance) coefficients to vary as a function of depth. In addition, the model allows one to compute the influences of submerged targets, bottom types or unique submerged targets or water column layers with defined by their reflectance signatures of unique absorption and backscatter characteristics. Model simulations are presented to demonstrate the utility of the model for development of remote sensing algorithms for use in coastal and marine water types.

Improvements to a layered analytical irradiance model for application to coastal waters with depth-dependent water constituents, various bottom types, and variable water depths

This paper compares homogeneous analytical two-flow equation solutions to an improved iterative solution technique of the same differential equations that describe the transfer of irradiance in a layered medium such as water with depth dependent water constituents. The layered model is developed to be used in oceanic or atmospheric models in order to provide a detailed mechanism for the influence of irradiant energy in the heating rate of the medium. The layered model is designed to generate synthetic water surface reflectance signatures in the presence of depth dependent water constituents, various bottom types, and variable water depths. In addition, the model allows one to compute the influences of submerged targets or layers with unique reflectance signatures or unique absorption and backscatter characteristics.

Modeling the influence of water waves upon remote sensing imagery: the underwater radiance distribution and shape factors

This paper describes the results of modeling the water wave surface and underwater light field as influenced by water waves using a Monte Carlo model (MCHSIM). Model and sensor data related to water column properties and benthic properties that influence the light upwelled from below the water - as observed from a sensor looking from below or above the water surface is presented. Synthetic image results using Monte Carlo techniques show the influence of water waves upon subsurface shape factors and these factors can be used in shallow water remote sensing algorithms that are based on underlying analytical models. The upwelling angular distribution of light is calculated from the model and results shown for 490 nm. The upwelling and downwelling shape factors are shown from model runs which compare the results with solar zenith angle for nadir viewing geometry, and for realistic water surface wave facets. It is clearly shown that shape factors are strongly dependent upon not only viewing geometry and zenith angle of the sun, but also upon water waves that can focus and defocus radiance entering a wind roughened water column and influence the shape factors due to the scattering lobe effect. This paper presents results quantifying the magnitude of water effects upon the upwelling and downwelling shape factors in a systematic and quantifiable manner at 490 nm and demonstrates the utility of the model to assess the influence of water waves in a full 3-D Monte Carlo hyperspectral synthetic image cube model that accounts for adjacency effects.

A sensitivity analysis of irradiance reflectance calculated from the modified two-flow equations with shape factors to various model inputs

A sensitivity analysis of the irradiance reflectance calculated from the modified two-flow equations (Bostater, et al., and Bostater et al, 2002) to various model inputs is addressed. The modified two-flow equation approximate of the radiative transfer equation (RTE) with a collimated or specular component is used to calculate the sensitivity of the calculated irradiance reflectance on the inclusion of solar zenith angle, wind speed, shape factors, chlorophyll-a concentration, water depth, and bottom type.

Synthetic image generation of shallow water using an iterative layered radiative transfer model with realistic water surface waves

Modeled hyperspectral reflectance signatures just above the water surface are obtained from an analytical radiative transport model applicable to shallow water types. Light transport within the water body is simulated using a fast, accurate radiative transfer model that calculates the light distribution in any layered media. A realistic water surface is synthesized using empirically-based statistical models of ocean surface waves. Images are displayed as 24 bit RGB images of the water surface using selected channels. The selected channels are centered at 480, 520 and 650 nm. Hyperspectral image cubes with two spatial and a third spectral dimension are shown to allow the detection of any optically unresolved features in the two-dimensional RGB image.

Data assimilation of AVHRR and MODIS data for land base initialization and boundary conditions in the UTC-M atmospheric boundary layer sea-breeze model of Space Coast Florida

The purpose of this paper is to present results of simulations of the Florida Tech UTC-M sea-breeze model with the addition of a simplified atmospheric downwelling radiation subroutine combined and a thermal inertia subroutine into the atmospheric planetary boundary layer model, in order to calculate time dependant heat flux boundary conditions at the air-land boundary that are derived from satellite data from AVHRR and MODIS sensors. The improved UTC-M planetary boundary layer model with this thermal sub-model subroutine is used to demonstrate the use of thermal inertia to help estimate heat fluxes at the land-air interface which in turn influences convergence and vertical fluxes near the bottom boundary, and which may affect mesoscale meteorological wind and seabreeze over complex land-water margins. Additionally, message passage interface (MPI) parallelizing Fortran techniques were used to improve the computational time when the model grid was decreased down to 2 or 1 km cell when simulations where performed on the FIT supercomputer based on an IBM Beowulf Linux cluster. We present some results of the UTC-M simulations and associated results due to the influence of the parameterization of the net surface radiation and thermal inertia using the spectral or wavelength (channel) specific data from MODIS and AVHRR satellite sensors.

Water and land surface satellite based data for use in the UTC-M Mesoscale planetary boundary layer atmospheric model: sea breeze predictions along central Florida using a thermal sub-model

The purpose of this paper is to present simulation results of a thermal sub-model developed for the Florida Tech UTC-M sea-breeze model. The insertion of this thermal radiative model into the atmospheric planetary boundary layer model, allows calculation of time dependant heat flux boundary conditions at the air-land boundary that are derived from satellite data such as AVHRR and MODIS. The improved UTC-M planetary boundary layer model with this thermal sub-model is used to demonstrate the use of thermal inertia to help estimate heat fluxes at the land-air interface which in turn influences convergence and vertical fluxes, which then affects mesoscale meteorological wind and convergence predictions. We present thermal radiative model simulations and associated results due to the influence of the parameterization of the net surface radiation and thermal inertia using wavelength or channel specific data from MODIS and AVHRR satellite sensors. Results are presented for cloudless sky conditions.

Sensitivity analysis of the Florida Tech UTC-M mesoscale atmopsheric seabreeze model to estimates of water and land heating rates using AVHRR SST and coastal margin land temperatures

In this paper we conduct sensitivity analyses of the land surface boundary conditions and parameterizations in the UTC-M primitive equation atmospheric planetary boundary layer Seabreeze model. The boundary conditions for temperature of the ocean and land surface is based upon satellite derived AVHRR estimates over the water and coastal land margins. The sensitivity analysis of the boundary conditions as well as the heating and cooling rates in the planetary boundary layer model are also described. The model domain of interest is the region over the Space Coats of Central Florida. This reign is unique because of its complex coastal water-land margin and its close proximity to the Gulf Stream and Cape Canaveral. This model study demonstrates a method whereby the ocean surface and atmosphere is couple using remotely sensed data for predicting the coastal Seabreeze and associated convective convergence and expected cloud development in the planetary boundary layer.