Scintillation indices as high as four to six have been measured in experiments in the atmosphere. The inner scale of turbulence has a dramatic effect on the value of the scintillation index when the index is greater than unity. Here we review theory which shows that, in order to obtain such high values, it is necessary to modify the Kolmogorov form by the addition of an inner scale. We present results for plane wave and point source initial conditions and show that, without an inner scale, the maximum scintillation index that may be obtained is much below experimental results. Moreover, we show that, even with the inner scale included, we must have a point-source-type initial condition rather than a plane-wave-type initial condition. We also discuss the effect of the particular modification of the Kolmogorov form that is used to introduce the inner scale.

A simulation was constructed and numerical experiments conducted of optical wave double pass propagation through a phase screen. The phase screen had a Gaussian phase fluctuation with a Kolmogorov power law spatial spectrum. The results showed a statistically partial phase conjugation.

A new analytical model of the spatial power spectrum of temperature fluctuations in the atmospheric boundary layer, showing the characteristic "bump" just prior to the dissipation ranges, is developed as a modification of the standard Tatarski model by taking into account the random fluctuations in the average dissipation rates er and xr Both er and xr are shown to satisfy the statistics of a gamma distribution and thus the joint distribution is assumed to be a bivariate gamma model, which shows good agreement with experimental data. Based on this new spectrum model, we then calculate the related structure function of temperature, showing deviations from the standard r2/3 power-law behavior of the Tatarski model. The refractive-index spectrum is assumed to have the same functional form as the temperature spectrum for optical propagation. Using it we also calculate the normalized variance of log-intensity for plane waves.

Successful design of an atmospheric laser communications terminal requires an understanding of atmospheric turbulence induced scintillation of the optical signal. Laser beam scintillation is small scale interference within the beam cross section due to turbulence induced fluctuations of the refractive index of the atmosphere, causing variations in the spatial power density at the receiver. The variations in the spatial power density at the receiver manifest themselves as fades and surges of the detected optical signal. By understanding the statistics and power spectrum of the fades and surges, communication terminals can be designed to achieve needed levels of performance by employing optimized choices of increased link margin and error coding. As part of the HAVE LACE (Laser Airborne Communications Experiment) program, amplitude scintillation data was collected and analyzed for extended propagation path lengths. The analysis included the determination of the statistics and temporal power spectrum of the scintillation and the effect on communications performance. Since the HAVE LACE terminals used direct detection of pulsed laser energy, the random variations in the received signal strength was used to evaluate only the atmospheric turbulence induced amplitude scintillations. The collected data has been reduced and compared with a model for extended path length channels. The objective of this comparison was to verify the performance of the model against the collected data. The results from the comparison show a reasonable degree of correlation between the data and the model which warrants further investigation of this approach. This analysis is presented in a form which is consistent with an understanding of the implications of the effect of the communications channel on system performance.

The intensity statistics for a laser beam propagating through a random phase screen was compared to the H-K distribution. Since the H-K distribution was derived from the conditionalization of the Rician distribution with the Gamma distribution, experimental intensity data was collected to verify these distributions. The intensity data was first lowpass filtered to yield statistical moments that were found to be in agreement with those of the gamma distribution. Next, the intensity data was segmented into small time ,intervals, in which all time segments with approximately the same variance were grouped together. The statistical moments for these new time series were then compared against the moments for the Rician distribution. The moments for the segmented time series matched the moments for the Rician distribution extremely well.

The experimental measurements of enhanced backscattering for a laser beam propagating through random screens is presented here. A 10 cm collimated laser beam was transmitted through a random phase screen and reflected back through the same screen by a mirror placed in the near field behind the screen. The laser beam was then imaged at infinity and the averaged scattered intensity was measured. Directly in the backscattered direction an enchancement peak was observed. The width of this peak was near the diffraction limit, while the height was about a factor of two larger than the normal large scattering element observed from phase screens.

We have been Concerned to ascertain just how well our understanding of white light speckle phenomenology matched the quantitative aspects of white light speckle imagery. For reasons of simplicity we have restricted our attention principally to the Labeyrie signal amplitude and the Knox-Thompson phasor per se, and to the image derivable there from. For this purpose we have processed in various ways two sets of short exposure, narrow spectral band "images." Each image is photodetection event shot noise limited in the sense that the image consists simply of the coordinates locating each of some random number of photodetection events. One set of short exposure "images" corresponded to the single star ζ-Delphinus, the other set to the binary star β-Delphinus.

Standard measures of image quality are used to quantify the effect of atmospheric turbulence on specific target and background signatures from digital images. The degraded signature corresponding to any atmospheric coherence diameter can be computed with a direct convolution model developed in previous work. Image analysis software has been developed to compute image quality factors by comparing the original and degraded signatures. Regions of maximum clutter can also be be located and evaluated with the image quality factors. Results are presented for signatures gathered with a FLIR imaging sensor under typical field conditions.

The current status of large aperture scintillation techniques for refractive index structure parameter measurement is reviewed, instrument design considerations and limitations are discussed, and a new incoherent aperture profiling system is described.

In order to characterize the effects of atmospheric turbulence on laser beam propagation, measurement campaigns based on optical techniques were organised on a 1500 meter-long near horizontal path.

Structure constants and inner scales of temperature fluctuations were derived from ultrasonic anemometer-thermometer measurements taken at heights of 48 m and 80 m above the ground. They were shown to follow local Monin-Oboukhov similarity from very unstable to very stable atmospheric stratification. A direct empirical expression for the stability dependence of the nondimensional inner scale is given. A bichromatic scintillometer based on the wavelengths 0.63 μm and 10.6 μm was operated near the ground. Comparison between optically measured path averaged inner scales to those derived from point measurement of vertical velocity fluctuations yielded excellent agreement. The observed dispersion of refractivity fluctuations was used to separate structure constants of temperature and humidity. Surface fluxes obtained from parameterizations requiring only simple meteorological input data were applied to scale optically measured structure constants of temperature and inner scale. The parameterized fluxes were shown to be sufficiently accurate to be used in Monin-Oboukhov similarity based models.

A method is proposed to determine the amplitude and phase fluctuations of a wave front at any particular point. The only information needed is a single output of a detector whose excitation is the integrated incoming and a reference wavefields.

Our perspective of Cn2 in low stratus clouds and their associated subcloud regions is developed in this paper. In so doing, two environments are considered, that of a rising parcel and that of the ambient environment, resulting in two vertical profiles of Cn2. An extension of Tatarski's formulations is used to characterize the ambient environment, assuming that the turbulent parcels are weakly conservative and passive; whereas, our definition of "optical turbulence" is used to address the rising parcel where conditions are not conservative and passive. Vertical profiles of n, dn/dz, and Cn2 associated with the rising parcel and those associated with the ambient atmosphere are presented for one set of conditions.

Measurements of the level of turbulence Cn2 have been successfully performed with the optical scintillometer. The success of this instrument is based on the observed fact that the variance of aperture-averaged scintillation is described by weak scattering theory even for conditions under which strong scintillation is observed for small apertures. However, for sufficiently long propagation paths the aperture-averaged variance is affected by strong scattering. The effects of strong scattering are calculated theoretically and compared to experiment. The physics of this regime are discussed and the important parameters investigated in order to extend the range of validity of optical scintillometer measurements.

The scintillation of starlight contains information about the refractive turbulence strength in the atmosphere. The fluctuations to each two-dimensional spatial wavenumber in the scintillation pattern are caused by turbulent features that have the same two-dimensional wavenumber. Therefore, a receiver that spatially filters the scintillations in starlight can measure the amount of turbulence in the atmosphere in a narrow band of wavenumbers. If the entire atmosphere were moving with constant velocity, the dominant wavenumber would produce a constant frequency as the turbulence moved across the filter. However, wind velocity typically varies with altitude, and turbulence at different altitudes will produce different frequencies. If the wind velocity profile is known, the vertical profile of Cn2 can be inferred from the frequency distribution of scintillations at a particular wavenumber.

Field tests have been performed on a prototype spatially-averaged filter scintillometer designed to measure wind components transverse to an optical path. Unlike previous scintillometers, the spatially-averaged filter design peaks weighting functions at five segments along the optical path, producing simultaneous cross-path wind readings from these path segments. Intercomparison tests indicate that the scintillometer performs well under most meteorological conditions. However, a performance problem was identified during cross-path wind reversals in light and variable winds, where ambiguities in the sign of the crossing wind component occasionally caused spurious wind readings. The calibration factor also needs further testing. The crosswind scintillometer has significant remote wind sensing applications.

The Target Contrast Characterizer (TCC) was described in detail by the authors last year in the Proceedings of SPIE Vol. 926. It consists of experimental equipment and methodology which permit images of near-field (or close up) and far-field (or at engagement range) targets and their backgrounds to be registered in near real-time for characterizing the effects of the atmosphere on inherent and propagated target contrast. This paper details the method for obtaining target contrast transmission measurements using the TCC. Measurements of target contrast transmission are compared with model predictions of atmospheric transmission. In addition, the very sensitive measure of pixel-to-pixel or area-to-area contrast transmission obtained with the TCC is discussed as well as improvements which should be added to optimize near-field/far-field image comparisons.

Broadband multispectral transmissometer systems operating in the visual through 14-μm band of the electromagnetic spectrum are carefully designed radiometers having narrow fields of view aimed toward temporally modulated sources. However, our analysis of transmissometer field test data has shown evidence that nonlinear system responses can often occur. In this paper, we present an analysis of the effects of nonlinear detector response on broadband transmissometer system response. Our results show, for example, that when the system detector responds nonlinearly, the transmittance data show effects due to background radiation that cannot be removed by presently used signal processing techniques. With the analysis, examples of broadband multispectral transmittance data are evaluated and characteristic features of data indicating evidence of nonlinear system response are identified.

This report describes the results of a theoretical investigation of the impacts of non-ideal transionospheric signal propagation on the imaging performance of space-based synthetic aperture radars (SARs). The nature and magnitude of the effects are demonstrated in quantitative terms versus the strength of the ionospheric disturbance. A number of different mitigation schemes for combatting the effects are also evaluated, including: optimal flight and viewing geometry, increased carrier frequency, and adaptive signal processing.

A newly-developed Atmospheric Turbulence Measurement and Observation System (ATMOS) was equipped with multiple apertures, allowing the phase structure function to be determined experimentally. Data was obtained with the ATMOS under two conditions: (a) over a horizontal path with a laser source, and (b) from near-vertical viewing of stellar sources. This data was compared to measurements of the refractive index structure parameter (Cn2) obtained from scintillometers, spatially separated temperature probes, and anemometers. The existence of the Kolmogorov spectrum of turbulence during the measurement period will be discussed as well as the relevance of the results on the transverse coherence length, r0. The optical arrangement, as well as data collection and analysis methods, are also outlined.

A prototype atmospheric turbulence 'seeing' monitor was upgraded to provide a means of freezing, capturing, and digitizing up to 35,000 frames of two-dimensional image data at rates up to 350 frames per second with submillisecond exposure times. Analysis of this data is presented, yielding crucial information about the frequency spectra of turbulence-induced angle-of-arrival fluctuations. This information is used to establish a range of measurement parameters which are suitable for characterizing the statistical nature of the fluctuations and to establish a practical means for experimentally determining periods of atmospheric stationarity.

Observations and theory for millimeter-wave propagation through clear-air turbulence, rain, fog, and snow are reviewed. Measurements have shown the effects of refractive and absorptive fluctuation in air. Measured quantities include the intensity, the phase difference between spaced antennas for a single electromagnetic frequency as well as phase difference at a single antenna for waves having differing frequencies. Typical statistics of these quantities are their variances, structure functions, temporal spectra, and probability distributions.

In this paper we describe recent enhancements to the RADTRAN algorithm including: (a) the capability to model polarized microwave surface emissivity for ocean, dry and moist soil, various types of vegetation, sea ice, and snow cover, (b) calculation of the scattering properties of precipitation, (c) treatment of multiple scattering by precipitation in the evaluation of background brightness temperature, and (d) the incorporation of simple statistical parameter retrieval methods.

Measurements of transmittance through falling snow were made in the visible, 3-5 μm and 8-12 μm bands along a 540 m path. In the visible, measurements were made for both a narrow (0.05 mrad) and wide (3mrad) transmitted beam detected by a common receiver. The apparent path averaged extinction, derived from the transmittance measurements using the Beer- Bouguer law, was found to be dependent upon wavelength and transmissometer geometry. The results may be explained by taking into account scattered contributions to the measured transmittance. The results are compared to calculations made with a single-scattering model and a multiscattering model.

Using atmospheric modulation contrast function area (mcfa) as a single-valued numerical criterion for image quality propagated through the atmosphere, a statistical study of atmospheric imaging data has led to the determination of regression coefficients with which to quantitatively predict at visible and near infrared wavelengths effects of windspeed, air temperature, and relative humidity on image quality propagated through the atmosphere as functions of wavelength and of spatial frequency. Utilization of this procedure is quite simple. One simply plugs in expected values for windspeed, air temperature, and relative humidity in the regression coefficient expression for mcfa. The larger the expected mcfa, the better the expected image quality. Models are presented for desert and non-desert atmospheres. Preliminary experimentation indicates the accuracy of the models is quite good and that the quality of image propagation through the atmosphere can be described as a simple function of some basic meteorological dependences previously unknown. These fundamental meteorolgical parameters are part of weather forecasts measured world-wide, are simple and inexpensive to measure and can be used to determine also overall atmospheric modulation contrast function as a function of spatial frequency and meteorological conditions. An important advantage of this new approach is its simplicity as well as accuracy.

Aerosol loadings and size distributions, covering the radius range from 0.08 to 23.5μm, have been measured at a coastal site on the island of South Uist, situated off north-west Scotland, for wind speeds ranging from essentially zero to values in excess of 30ms-1. The dependence of aerosol volumetric loading upon wind speed indicate that these loadings continue to increase with increasing wind speed up to the limit of the data set, in contrast to earlier suggestions of a levelling off for speeds beyond about 14ms-1. The temporal decline of particulate concentrations at lower wind speeds is consistent with the results of a simple turbulent deposition model. For a given wind speed, estimates of atmospheric extinction due to the aerosol particles show little variation from visible to infra-red wavelengths. It is suggested that these findings are a consequence of the domination of extinction by the larger aerosol particles found in this environment.

A model is developed to calculate the vertical variation of aerosol extinction coefficients throughout the marine atmospheric boundary layer. It is a mixture of empirical and physical models, formulated to describe the often observed non-uniform, but also non-logarithmic, profiles. The physical model is based on the dynamical processes affecting the production, mixing, deposition and size of the aerosol within the marine atmosphere. A status report is presented including a critical evaluation.

Low and high resolution measurements (20 and 1 cm-1) of atmospheric transmittance between 1800 and 3500 cm-1 over a 5.7 km horizontal path under extreme conditions of temperature are reported. Results are compared to calculations using the LOWTRAN 6 and FASCOD2 transmission codes. In particular, we examine the accuracies with which these codes predict transmittance in spectral domains (1800-2000 and 3200-3500 cm-1) strongly affected by water vapour concentration. Preliminary analysis indicates that, although LOWTRAN 6 predicts well the summer transmittance (+30.3 °C) there are significant errors for the winter case (-21.4 °C). In this case, LOWTRAN 6 underestimates the transmittance by about 20 % in the region near 2000 cm-1 and 3300 cm-1. The comparison of FASCOD2 with experimental results shows a much closer agreement than for LOWTRAN 6 for both summer and winter cases.

One goal of system performance modeling is to predict system performance for new, untested environments. When the relationships between environmental variables and system performance variables are clear, the solution reduces to an uncomplicated sensitivity analysis. For complex systems, however, such as a sequence of preprocessing, detection and classification algorithms in an Aided Target Recognition box, the environmental relationships may not be obvious. This paper explores two approaches to including the atmosphere in extrapolating imaging system performance in different environments. One approach is to describe analytically the response of individual image processing steps to atmospheric effects. We consider the sensitivity of an edge detector, target moments, and a basic linear mapping classifier to obscuration, path radiance and turbulence blur. In a second approach one modifies the input images (for example the training images for a classifier) for different weather conditions and then observes the system response to these new inputs. The problems of image modification that remove one set of atmospheric effects while adding new conditions are also discussed.

Thermodynamics, turbulence, and dynamics of damp, saturated and supersaturated air are systematically considered to produce a unified low stratus cloud-subcloud microphysics model. The model produces vertical profiles of temperature, pressure, density, specific humidity, relative humidity, liquid water content, drop size distributions, wind and stability based on inputs of temperature, pressure, relative humidity, visibility, and wind at a reference height (2 m), the vertical ascent rate at the reference height and the top of the cloud, and the height of the top of the cloud.

Drop size distributions from a composite of models are used to generate vertical profiles of extinction, backscatter, and absorption in low stratus clouds and associated subcloud regions. Interpolation by splines is used in estimating total cross-sectional areas of droplets in relatively narrow intervals to produce relatively smooth profiles. Drop size distributions and vertical profiles of effective complex indices of refraction of drops, extinction, backscatter, scattering, absorption, and meteorological range for a wavelength of 0.55 μm are presented.

The alexandrite laser is a flashlamp pumped, solid state pulsed laser that utilizes the alexandrite crystal for the lasing medium with output radiation between 0.755 and 0.760 micron wavelengths. Since the alexandrite laser is relatively new, little work has been done to study the propagation effects at this wavelength. The U.S. Army Missile Command at Redstone Arsenal, Alabama recently acquired two alexandrite lasers which have been used to study the effects of this radiation under different conditions, including atmospheric propagation. Baseline information was gathered in the laboratory for each laser prior to propagation studies at range. The lasers were characterized in terms of pulse length, energy, pulse shape, and pulse-to-pulse energy variation. Energy measurements were taken with a special device at 200 and 400 meters to determine the intensity profile of the beam. Energy measurements were taken at 1200 meters with a 1 cm2 detector, positioned in the approximate center of the beam. At the same time a 3-inch diameter section of the beam was sampled by using a beam splitter and lens to focus that portion into a radiometer. Beam spot shape and location at the target were observed. Statistical techniques including frequency distribution curves of fluence in the two sample areas were used to analyze the fluence at 1200 meters and the effects of atmospheric conditions, especially scintillation, on the transmitted energy. Atmospheric data was recorded by an Atmospheric Sciences Laboratory Meteorological Station at the range. A scintillometer located adjacent to the beam path continuously recorded Cn2 values during the test program.

A 500m experimental propagation range has been implemented, with transmission systems operating at 37, 57, 97 and 137 GHz in the millimetre-wavelength region of the electromagnetic spectrum and at 10.6 µm and 0.63 µm in the infrared and optical regions, respectively. A comprehensive range of meteorological measurements complements the propagation data, which together are used to carry out a variety of investigations into the interactions between electromagnetic radiation and the atmosphere. Of particular interest are the effects of precipitation, including rain, snow and hail, hydrometeors such as fog, and other atmospheric phenomena including scintillation. The results are further used to compile statistical data to evaluate the reliability of future communications systems operating at millimetre wavelengths.

In order to evaluate the series expansion in scattering orders derived by Tam and Zardecki for the multiple forward scattering irradiance, the multidimensional integrals describing the contributions of the individual scattering orders n have been transformed into 1-D ones by use of a weight function Fn. The function Fn represents that section of a spherical hypersurface centered at the origin which is enclosed within the unit hypercube. For the calculation of the Fn, two approaches are proposed. The first one starts from a combinatorial consideration and yields a complete mathematical expression for the Fn in the form of multidimensional integrals which, however, can be computed recursively from one another in order of increasing n by a 1-D analytical or numerical integration. For higher n a second approach is developed yielding for the Fn an approximation in form of a function series which is used to expand the contributions of the individual scattering orders in a fast converging series in negative powers of n. This expansion also reveals general features of multiple forward scattering.

To study the response of a medium to a localized disturbance, the coupled aerosol-beam equations-in which the dominant interactions are diffusive mass transport and conductive energy transport-are solved numerically, thus giving the spatio-temporal behavior of the propagating beam and the irradiated aerosols. In the context of the cloud clearing problem, a deleterious effect of recondensation is assessed. We examine the effect of turbulence on the distribution of droplet sizes during the recondensation. Results relevant to propagation and imaging are given.

Scattering of electromagnetic millimeter waves by snow is represented by Rayleigh-Gans theory applied to a polydispersion of equivalent particles that are circular disks. The circular disks are assumed to be randomly oriented. The material that comprises the equivalent particles consists of a homogeneous mixture of ice and air that produces an effective refractive index that is represented by the Bruggemann mixing rule. The fraction of ice in the equivalent particle varies as the size of the largest dimension; it is .35 for the smaller particles and it diminishes as the largest dimension increases. Extinction coefficients, phase functions, asymmetry factors and backscattering cross sections are presented for frequencies 30 < v < 225 GHz. Comparisons with available measurements are presented.