Absolute radiometric calibrations of Landsat 5 Thematic Mapper satellite are improved with the inclusion of a method to invert optical depth measurements to obtain an aerosol particle size distribution and a non-lambertian surface reflectance model. Also, a program is developed to improve speed and standardization of the entire calibration procedure. The inverted size distributions can predict radiances varying from the previous jungian distributions by as much as 5 percent, though the reduction in the estimated error is less than one percent. An empirical model for the surface reflection of White Sands using a two-degree polynomial fit as a function of scattering angle was employed. The model reduced estimated errors in radiance predictions by up to one percent. Satellite calibrations dating from October, 1984 are reprocessed using the improved methods and a linear estimation of satellite counts per unit radiance versus time since launch is determined.

The light scattered from a scatterer depends on the geometrical properties such as size, shape, and their distributions as well as electromagnetic properties such as the complex index of reflection. The four major Mueller scattering matrix elements have been experimentally measured for an aluminum rough surface scattering laser light at lambda = 441.6nm for various incident angles. Measurements were also made for non-conducting diffuse surface and an aluminum coated diffuse surface. The sixteen Mueller matrix elements of these diffuse surfaces were measured in order to study the relative role of reflectance and roughness for scattering from a rough surface. Some representative matrix elements for the rough surfaces as well as for conducting and non-conducting rough surfaces are shown. S11 and S34 are sensitive to illumination angles and surface variation. S33 are sensitive to illumination angles only. S34 are very sensitive to the change of refractive index from real to complex.

The design, fabrication, and testing of diffractive microlens arrays for use in fiber optic interconnects is presented. Advantages of using diffractive microlenses for fiber interconnects instead of refractive microlenses are discussed. A theoretical discussion of diffractive lens operation including the effects of chromatic aberration on fiber coupling is given, along with equations for mask generation. From these equations, a set of masks to fabricate an array of four phase level diffractive microlenses was produced. Experimental procedures for the fabrication of diffractive microlenses from this mask set are presented. Arrays of binary lenses in photoresist with depth errors of less than 4% were fabricated and test results are given. A novel fiber optic tap utilizing both surface relief microlenses and a volume holographic element is demonstrated. In addition, a fiber array coupler using silicon V-grooves and an array of diffractive lenses is also demonstrated.

No description available.

<p> Efficient nonlinear optical devices are designed and demonstrated in "photonic molecule''-like coupled-cavity resonator systems on a semiconductor chip. A coupled-cavity resonator may be designed to support distributed supermodes, and to allow independent control of the resonant frequency and linewidth of each supermode. Such control allows reduction of dispersion without compromising effective nonlinearity in the resonator, as well as the design of anisotropic output coupling or radiation that allows optimized nonlinear functions. Therefore this resonator manifests itself as a favorable platform for building nonlinear devices including optical parametric wavelength converters and oscillators based on four-wave mixing that call for different couplings to the signal, pump and idler modes. A physical model based on coupled-mode theory describes all relevant linear and nonlinear processes in triply-resonant microcavities, and a generalization of the usual nonlinear figure of merit is proposed to evaluate the effects of distributed supermodes on nonlinear conversion efficiency in such devices. Experimental work is presented that demonstrates coupled cavity devices for wavelength conversion in crystalline silicon, where two-photon absorption sets conversion efficiency limitations. In addition, an investigation of deposition conditions of hydrogenated amorphous silicon is described where amorphous silicon allows for a higher nonlinear figure of merit than crystalline silicon, promising increased performance in such devices. More generally, mode interference and coupling in coupled-cavity resonators, as a unique degree of freedom in integrated optics, is explored through designs of linear devices including efficient optical filters, wavelength converters, and modulators. </p>

QC 351 A7 no. 28 / A simple functional relationship was found between normalized photographic resolving power (RN) and target modulation (M) given by RN = Mn, where RN is the ratio of the resolving power at a modulation (M) to the resolving power for M > 0.98 and n is an arbitrary constant. Since this equation provides a useful approximation to most of the measured curves, the resolving power dependence on target modulation for a given lens -film combination can be specified by the high contrast AWAR and the constant n. The value of n was found to lie between 0.46 and 0.62 for the AWAR versus target modulation curves of the 10 lens -film combinations analyzed. The effects of such factors as development, filter, and exposure on this relationship were found to be generally weak and seldom simply predictable. In tests to determine typical values of resolving power reading differences, an 8% average difference in photographic resolving power was encountered over 240 readings by three observers on Plus -X Aerecon film. A comparison of photographic resolving power versus target modulation curves obtained from experiments using seven test targets with those obtained from experiments using fewer targets showed that four targets, with modulations of approximately 0.11, 0.29, 0.73, and 0.98 +, were sufficient to determine the resolving power dependence for M > 0.05.

QC 351 A7 no. 64 / Generalized bending is a one -parameter family of changes to two curvatures and related thicknesses of a previously defined optical system consisting of spherical and plane refracting surfaces. This family of changes leaves first-order properties invariant at all other surfaces in the system. Thus, third-order aberrations at the other surfaces are also unchanged. The third-order aberrations may then be expressed as functions of independent generalized bends at different locations; therefore, simultaneous correction of several aberrations is possible. Comparison of ray fan plots for real rays through an optical system shows marked differences for various degrees and locations of generalized bending. Surfaces at which a generalized bend would make significant changes to aberrations of the original lens are easily identified. This use of generalized bending would be helpful in advanced stages of a design routine.

QC 351 A7 no. 80 / This dissertation presents a theoretical investigation into the production, evolution, and asymptotic form of chirped pulses in homogeneously and inhomogeneously broadened laser amplifiers. Amplifier equations of motion are obtained in a form appropriate for dealing with such frequency-modulated pulses. The transient response of laser amplifiers to variously chirped Gaussian input pulses is studied parametrically using numerical solutions of the amplifier equations. The chirping mechanisms of the intensity dependence (Kerr effect) and the quadratic frequency dependence of the index of refraction are discussed briefly, as are the chirps produced by them and the amplification of Gaussian pulses in their presence. The amplifier whose host exhibits these dispersive effects is treated as a sequence of pairs of slices. One of each pair amplifies and exhibits the Kerr effect; the other has an index with a quadratic frequency dependence. This slice model is used to obtain steadystate pulses in both homogeneously and inhomogeneously broadened amplifiers whose host indexes have a quadratic frequency dependence. The steady-state pulse characteristics are determined as functions of amplifier parameters and the index curvature. The principal results are as follows: The homogeneously broadened amplifier responds predominately to the temporal character of a chirped input pulse whereas the inhomogeneously broadened amplifier response depends primarily upon the pulse spectrum. Of three important concepts (area theorem, echoes, and optical nutation) used to describe unchirped pulse amplification in inhomogeneously broadened media, only photon echo is useful when pulses are more than slightly chirped. The presence of the Kerr effect can produce significant chirps on large pulses. Amplification in the presence of the Kerr effect produces pulses strikingly similar to experimental results. Quadratic frequency dependence in the index has very little influence on most pulses in short amplifiers but has a cumulative effect in long amplifiers and laser oscillators. Chirped steady-state pulses exist in both homogeneously and inhomogeneously broadened amplifiers when the host index has such frequency dependence. In the homogeneously broadened case, they exist at relative gain levels dramatically below other theoretical predictions. They occur in the inhomogeneously broadened case only for the smaller index curvatures.

The generation of single CO2 laser pulses, having sub-nanosecond rise and fall time, is described. The duration of these pulses is controlled in the range 0.83 to 100ns and their peak power is of the order of 1 MW. For the generation of these pulses, a double-discharge TEA CO2 laser has been used, in conjunction with an electro-optical shutter. A telescopic CO2 laser amplifier has been employed to amplify the nanosecond pulses to 1 GW power levels. The design and operational characteristics of this device are presented and the effective elimination of associated problems is discussed. The effect of organic vapour additives on the performance of this oscillator-amplifier system is investigated. Experimental evidence is presented indicating an increased efficiency in the amplification of multi-line - as opposed to single line - nanosecond pulses. A novel method for single longitudinal mode operation of pulsed TEA CO2 lasers - incorporating the use of two unstable coupled resonators - has been developed. The powers thereby achieved are in excess of 30 MW.

A study of the performance of high energy coaxial flashtube systems designed for pumping fluid laser media is presented. The properties of such systems are described and construction details of the flashtubes studied are given. Temporal and spectral characteristics of the optical output from the flashtubes have been measured at various values of system parameters. Optimum system parameters for high luminous efficiency, uniform discharge pattern and long life of the flashtubes have been obtained. The effects of radio-frequency preionisation on the optical output, discharge pattern and overall performance of coaxial flashtubes have been investigated. The use of these coaxial flashtubes as potential pump sources for organic dye and iodine photo dissociation lasers is considered. Efficient laser output from atomic iodine is reported.