Engineering topics
Publications and source records attributed to Pearce, W. A..
Monte Carlo radiative transfer simulations are used to study the atmospheric spread function appropriate to satellite-based sensing of the earth's surface. The parameters which are explored include the nadir angle of view, the size distribution of the atmospheric aerosol, and the aerosol vertical profile.
On relatively clear days, the shadow cast by an isolated, optically dense cloud produces an effect which is similar in extent and magnitude to the adjacency effects associated with high-contrast ground albedo boundaries. A Monte Carlo radiative transfer model has been used to examine the intensity and its components, comparing the cloud shadow case with that for a Lambertian albedo boundary. Although the behavior of the total intensity is nearly the same for both cases, the behaviors of the intensity components differ significantly.
Improvements in remote sensing capabilities hinge very directly upon attaining an understanding of the physical processes contributing to the measurements. In order to devise new measurement strategies and to learn better techniques for processing remotely gathered data, it is necessary to understand and to characterize the complex radiative interactions of the atmosphere-surface system. In particular, it is important to understand the role of atmospheric structure, ground reflectance inhomogeneity and ground bidirectional reflectance type. The goals, then, are to model, analyze, and parameterize the observable effects of three dimensional atmospheric structure and composition and two dimensional variations in ground albedo and bidirectional reflectance. To achieve these goals, a Monte Carlo radiative transfer code is employed to model and analyze the effects of many of the complications which are present in nature.
The use of Monte Carlo radiative transfer codes to simulate the effects on remote sensing in visible and infrared wavelengths of variables which affect classification is examined. These variables include detector viewing angle, atmospheric aerosol size distribution, aerosol vertical and horizontal distribution (e.g., finite clouds), the form of the bidirectional ground reflectance function, and horizontal variability of reflectance type and reflectivity (albedo). These simulations are used to characterize the sensitivity of observables (intensity and polarization) to variations in the underlying physical parameters both to improve algorithms for the removal of atmospheric effects and to identify techniques which can improve classification accuracy. It was necessary to revise and validate the simulation codes (CTRANS, ARTRAN, and the Mie scattering code) to improve efficiency and accommodate a new operational environment, and to build the basic software tools for acquisition and off-line manipulation of simulation results. Initial calculations compare cases in which increasing amounts of aerosol are shifted into the stratosphere, maintaining a constant optical depth. In the case of moderate aerosol optical depth, the effect on the spread function is to scale it linearly as would be expected from a single scattering model. Varying the viewing angle appears to provide the same qualitative effect as modifying the vertical optical depth (for Lambertian ground reflectance).
Satellite instruments such as the backscattered ultraviolet (BUV) apparatus, have been in operation for a number of years. One current difficulty is the validation of the ozone inferences obtained from the BUV measurements using independent instruments. For higher altitudes rocket instruments are necessary. Two instruments currently under development include a chemiluminescent detector described by Hilsenrath et al. (1969) and a filter photometer rocket ozonesonde (Rocoz) developed by Krueger and McBridge (1968). The present investigation is concerned with an analysis of the Rocoz system, the information content of the measurements, and the utility of the system for intercomparison with the BUV system. It is found that the sampling characteristics of the Rocoz and BUV systems exhibit some fundamental differences. However, their results can be related through knowledge of the relation between pressure and altitude. This is best obtained through the solution of the hypsometric equation using rocket temperature measurements.
The refraction effect from three perspectives was considered. An analysis of the axioms on which the accepted correction algorithms were based was the first priority. The integrity of the meteorological measurements on which the correction model is based was also considered and a large quantity of laser observations was processed in an effort to detect any serious anomalies in them. The effect of refraction errors on geodetic parameters estimated from laser data using the most recent analysis procedures was the focus of the third element of study. The results concentrate on refraction errors which were found to be critical in the eventual use of the data for measurements of crustal dynamics.
The paper presents the results of a Monte Carlo simulation study of the brightness and polarization at right angles to the solar direction both for ground-based observations (looking up) and for satellite-based systems (looking down). Calculations have been made for a solar zenith angle whose cosine was 0.6 and wavelengths ranging from 3500 A to 9500 A. A sensitivity of signatures to total aerosol loading, aerosol particle size distribution and refractive index, and the surface reflectance albedo has been demonstrated. For Lambertian-type surface reflection the albedo effects enter solely through the intensity sensitivity, and very high correlations have been found between the polarization term signatures for the ground-based and satellite-based systems. Potential applications of these results for local albedo predictions and satellite imaging systems recalibrations are discussed.