Discussions are presented on laser-induced plasma generation in free air propagation (no target surface present). Plasmas can be generated with very high flux levels (for clean air) or with much lower flux levels (for dust-laden air). Pure water aerosols have not proven to substantially reduce air breakdown thresholds below those for clean air. Clean air breakdown at laser wavelengths has been predicted successfully for many years using theories developed for microwave wavelengths. By directly extending those theories to laser wavelengths, the dependencies on ambient pressure, spot size, wavelength and intensities have been predicted with favorable comparison to experimental data. Pure water aerosols (fogs, rain, clouds) have been investigated as to the phenomenology to be expected when irradiated by pulsed and CW laser devices at CO2 and DF laser wave-lengths. These aerosols are shown to heat, vaporize and/or shatter at various incident flux and fluence levels. The data base appears to substantiate this phenomology, but no substantive reduction of clean air breakdown thresholds has been observed when water aerosols are present. Dry solid single-material aerosols (e.g., Si0₂ dust) have been examined, both analytically and experimentally, for verification of phenomenology at various flux and fluence levels. Air breakdown induced by the onset of substantial vaporization rates of the irradiated particulate has been shown to be coincident with the onset of dirty air break-down. Real-world and man-made aerosol clouds have not been studied extensively, if at all, to the author's knowledge. Of particular importance are the effects on high energy laser beams of dust, vehicle exhaust, smoke, road dust and maritime aerosols (e.g., sea spray, sea fogs, etc.). For these important components of the HEL propagation story, no substantial HEL propagation data base exists.

The effects of molecular absorption, molecular line shape, continuum absorption, and aerosol extinction on laser radiation propagating through the earth's atmosphere will be presented. Model predictions of atmospheric attenuation of selected laser frequencies will be shown as a function of atmospheric conditions and slant path. The spectral region from visible to mm waves will be considered.

We have investigated atmospheric absorption near the frequency of the chemical oxygen iodine laser (COIL). Survey spectra were taken around the laser line both with a Fourier Transform Spectrometer (FTS) and with a tunable laser; we found the FTS to be better suited for this use. We have also measured absorption cross sections of several atmospheric gases at the COIL frequency. The photoacoustic detector used with COIL is preferred over the FTS.

Differential absorption lidar techniques for determining basic atmospheric properties to high accuracy are described. The theory of a two-wavelength lidar temperature measurement is given with simulation results that show accuracies of 0.5 to 1 K for measurements from ground-based and satellite platforms. The theory of measurements of the atmospheric pressure profile using absorption troughs, broad regions of nearly uniform absorption, is also described with simulation results that show accuracies of 0.1 to 0.3 percent. Recent measurements which confirm the accuracy of these techniques are reviewed.

Single-ended atmospheric transmission measurements employing laser radiation backscattered from a remote target are subject to uncertainties due to large fluctuations of the backscattered signals. The magnitude of these fluctuations is dependent upon atmospheric effects, the type of ba.ckscattering target (specular or diffuse), and the nature of the detection system employed (heterodyne or direct). The measurement uncertainty is reduced (i. e. , the signal-to-noise ratio is increased) by signal averaging over a large number of pulses. For signal averaging over n independent measurermts, the standard deviation of the average value, and hence the measurement uncertainty, is expected to decrease as n^-1/2. However, it has been found that both short-term and long-term changes in the atmosphere give rise to temporal correlation among successive laser return signals, and this correlation can significantly limit the improvement signal averaging makes in the accuracy of atmospheric transmission measurements. In this paper we give the results of several of our experiments which directly measured the limitations of signal averaging in reducing the standard deviation of laser signals backscattered from a specular or a diffusely reflecting target using both heterodyne- and direct-detection systems. These limitations are shown to be in agreement with a theory which takes the observed atmospherically induced temporal correlation into account.

By neglecting one of the transverse coordinates in the computer analysis of laser propagation one can study problems requiring high spatial resolution or many realizations of the atmosphere in a cost-effective way. A one-transverse dimensional wave optics code treating diffraction, turbulence and thermal blooming is described. Results are presented relating this approximate approach to more exact codes with full spatial dimensionality.

This paper reviews the theoretical formulations for exact and approximate methods based on the small-angle approximation solutions of the radiative transfer equations for laser beam propagation in dense spherical particles; for example, fog and clouds. Numerical results will be given for multiple scattering effects as functions of optical depth, detector field of view and detector's aperture size for typical cases of clouds. Results will also be given for contributions from the various orders of scattering.

A universal math model has been developed for the irradiance fluctuations of an optical beam propagating through atmospheric turbulence. This new model was developed under the assumption that the field irradiance consists of two principal compcnents, each of which has an amplitude that is m distributed. By comparing this model with our own experimental data, as well as comparable data of other researchers, we found both qualitative and quantitative agreement over all ranges tested and strengths of turbulence. The statistics used in making these comparisons were the first five normalized moments of the been' intensity. Our own data were obtained over a very homogeneous environment with ranges from 180 m to 3 km. The data so obtained correspond to many conditions of turbulence, from weak where lognormal statistics are ordinarily valid, to superstrong turbulence where negative exponential statistics exist in the limit. In the weak turbulence regime we found that the data support the lognormal and universal model, which both predict essentially the same results. For conditions of moderate to strong turbulence, both the K distribution and the universal model match the data. However, for very strong turbulence prior to the limiting form of the negative exponential distribution, only our universal model matched the data. One of the interesting observations that we made in our data was that the normalized moments follow a looped path in the vicinity of their maximum values when they are plotted as functions of the normalized second moment. None of the standard theoretical models (i.e., lognormal, K, and negative exponential distributions) predict this looping effect since they are only two-parameter models whereas our universal model is a three-parameter model.

The definition of the inner scale for atmospheric turbulence is given. Methods of measuring inner scale are discussed and typical values are given. Innerscale effects on irra-diance covariance and log-amplitude covariance are discussed for both weak and strong scintillation. The covariance width is limited by the inner scale for saturated scintillation.

We have developed a closed-form expression for the intensity variance of a spatially partially coherent source with a finite aperture. For the partially coherent source pair-correlated field statistics and Gaussian-phase statistics are introduced and the results compared. It is found that pair-correlated field statistics for the source are in good agreement with expected results for coherent and incoherent sources, while Gaussian phase statistics give unreasonably high scintillation results for incoherent sources.

Numerical solutions of the fourth moment differential equation are obtained for a two-dimensional homogeneous and isotropic random medium which is characterized by a Gaussian correlation function. In addition to the covariance of the intensity fluctuations, the full spatial dependence of the fourth moment of the propagating field is described for both plane waves as well as for finite beams. Results are also presented for the interesting geometry in which the four observation points do not form a parallelogram. In the case of an initially Gaussian beam, the dependence of the structure of the fourth moment on the beam diameter is investigated for several propagation distances. The results for the intensity fluctuations index σ21 , are compared with various formulations of the extended Huygens-Fresnel principle.

Atmospheric turbulence and target induced speckle can have a significant and deleterious effect on the performance of optical systems. The primary effect is to introduce a strong fluctuation in the intensity and a random perturbation of the phase of the fields at the optical receiver. This is equivalent to introducing a large noise source at the receiver; and consequently multiple samples and/or clever processing techniques are needed to accurately estimate the magnitude of the received intensity. In addition, for coherent detection schemes, the random phase and finite transverse coherence length introduced by the speckle and turbulence will limit the size of aperture that can be used. These characteristics of the speckle field after propagation through turbulence and its effect on optical systems will be discussed in a tutorial overview. In addition, recent advances in understanding and characterizing this phenomena will be presented.

Atmospheric electromagnetic wave propagation plays a crucial role in determining the performance of communication and radar systems that operate at millimeter through visible wave-lengths. Theoretical treatments of atmospheric wave propagation draw upon work on the micrometeorology of the medium, and feed work on the design and performance analysis of communications and radar systems. This paper will illustrate some of the features of the foregoing hierarchy, focusing on the interaction between the propagation and system analyses. Specific examples include: optical communication through atmospheric turbulence, coherent laser radar operation in atmospheric turbulence, and millimeter-wave communication through rain.

The effect of multiple scattering on the validity of the Beer-Lambert law is discussed for a wide range of particle size parameters and optical depths. To predict the amount of received radiant power, appropriate correction terms are introduced. For particles larger than or comparable to the wavelength of radiation the small-angle approximation is adequate, whereas for small densely packed particles the diffusion theory is advantageously employed. These two approaches are used in the context of the problem of laser beam propagation in a dense aerosol medium. In addition, preliminary results obtained by using a two-dimensional finite element discrete-ordinates transport code are described. Multiple scattering effects for laser propagation in fog, cloud, rain and aerosol cloud are modeled.

This paper reviews the application of radiative transport theory to the characterization of optical propagation in the marine channel.

A general model that describes beam wave propagation through a dispersing and absorbing, clear, turbulent, heterogeneous atmosphere is extended and related to common experimental quantities, viz, intensity covariances, angle-of-arrival, and the mutual coherence function. These experimental quantities are assumed to be measured via parabolic reflector antennas; aperture averaging is taken into account. Explicit expressions for the second and fourth order coherence functions for a general beam wave are derived; it is found that log-amplitude! Phase correlations play an important part in the derivation of these two auantities. A brief review of reflector antenna theory is given. Three experimental situations are considered: the quasi-optical method; intensity covariance via two reflector antennas; the long base-line interferometric method.

Atmospheric turbulence effects on laser beam propagation have been limited to the cases of very short exposure in the near-field and far-field and to the case of very long exposure. However for typical values of wavelength, aperture diameter, and pathlength, the Fresnel number can be on the order of unity. Thus neither the near- nor far-field condition is appropriate. In addition, pulse duration times can range from nonoseconds to seconds, thus neither the short nor long exposure limit is, in general, applicable. Since the limiting conditions do not apply to typical scenarios, a general interpolation expression for laser beam spreading, valid for arbitrary values of the Fresnel number and pulse duration, has been developed. This expression gives the proper limiting results in the near- and far-field, short exposure cases and long exposure case and entirely reasonable results in the intermediate cases.

The deployment of high energy lasers for endoatmospheric applications has been seriously hampered by the deleterious effects of thermal blooming on laser beam propagation. Early attempts to overcome these limitations involved altering the intensity distribution and/or obscuration of the exit aperture, a technique known as aperture intensity profile tailoring. A more promising approach to remedy propagation distortions has been the use of coherent optical adaptive techniques (COAT). Although this approach will compensate for turbulence as well, there are several problems involved in utilizing this technique for severe thermal blooming environments. These difficulties include performance improvement limitations due to near-target distortions, multi-glint sensing problems, and cost/complexity considerations. The apparent lack of a cost/performance effective approach to thermal blooming control led to this investigation of a multiple aperture construct (MAC) approach. The concept employs small discrete subapertures in place of a large single aperture to reduce thermal blooming effects in the target region. Preliminary calculations indicate that subunit phase control is not essential to achieving adequate performance. Wave-optics computer codes were used to study key performance parameters in addition to predicting target irradiances for several simple system models. The implication of this initial study is that multiaperture systems can be configured to achieve comparable or improved (~ 30%) performance levels as monolithic systems, but with smaller less costly optical elements.

The first portion of this paper addresses the impact of thermal blooming in viewing extended target images. Topics discussed include the nature of the degradation and how images are modeled based on principles of incoherent imaging and anisoplanatism. Characteristics of images for several realistic scenarios are discussed, quantitatively, in the context of parameters normally associated with the forward propagation, only. The wave optics approach used in the simulation is shown to agree with published experimental results. The second portion of the discussion deals with image compensation in correcting for the effects of blooming. Such compensation is achieved by simulating a multiple thermal source, low bandwidth, return wave adaptive optics system. Incoherent point spread functions are corrected through various Zernike modes based on information from the closed loop optimization. Each spread function, corresponding to an isoplanatic region, is convolved with the appropriate object field to reconstruct the improved, extended image. Residual image distortion, degradation in peak irradiance, and adaptive optics loop stability is discussed for several examples, with respect to degree of correction and wavelength sensitivity.*

High energy laser systems usually attempt to focus a small spot of laser radiation on a distant receiver plane. Aberrations within the optical system, as well as in the intervening atmosphere (turbulence), tend to spread the size of the beam in the focal plane and reduce the peak irradiance that could otherwise be obtained. Adaptive optics technology in wavefront sensors and deformable mirrors can significantly reduce the phase errors on the laser beam and increase the laser fluence on the receiver plane. The key to making this work is being able to measure the optical distortions accumulated along the actual laser beam path. To obtain a wavefront on which to make this measurement, a wavefront sensor must optically share the full aperture of the optical path with the outgoing laser beam.

The temperature and OPD's are found for the time dependent conduction equation with axial convection for an arbitrary heat source. The solution satisfies Dirichlet boundary conditions in the radial and axial directions.

Normally one of the two approaches is used to evaluate the performance of an adaptive optics system. In the first approach the normalized antenna gain (or Strehl ratio) associated with a variety of degrading effects acting alone is evaluated. The normalized antenna gain of the system when degraded by a combination of these effects is then found by forming the product of the individual antenna gains involved. The second approach is to evaluate system performance by simulation on a large mainframe computer. In the work presented here, physical optics formulas and elementary statistical concepts are used to develop an approach that shares some of the advantages of both of these previous approaches. By working in the time domain, relatively simple formulas are developed that shed insight into the factors that degrade adaptive optics system performance. In addition, the impact of several degrading factors acting simultaneously can be evaluated. In this study the normalized antenna gain is evaluated for a system degraded by turbulence, anisoplanatism, a finite servo bandwidth and a combination of anisoplanatism and a finite servo bandwidth. In the finite servo bandwidth study, system performance is evaluated as a function of diameter illustrating that the resulting antenna gain decays from the diffraction limited antenna gain to the large diameter asymptotic limit antenna gain first predicted by Greenwood. The diameter dependence of this effect is similar to that due to anisoplanatism but not as severe at intermediate diameters. The last study evaluating the combined degradation associated with anisoplanatism and a finite servo bandwidth illustrates the important role played by high temporal frequency phase information. Introducing a finite bandwidth to a system already degraded by anisoplanatism can actually improve performance slightly, in certain cases, because highly distorted high frequency information is lost.