The Microwave Remote Sensing Laboratory (MIRSL) st the University of Massachusetts has developed a unique single antenna, dual-frequency polarimetric Cloud Profiling Radar System (CPRS). This project was funded by the Department of Energy's Atmospheric Radiation Measurement (ARM) program, and was intended to help fill the void of ground-based remote sensors capable of characterizing cloud microphysical properties. CPRS is unique in that it can simultaneously measure the complex power backscattered from clouds at 33 GHz and 95 GHz through the same aperture. Both the 33 GHz and 95 GHz channels can transmit pulse-to-pulse selectable vertical or horizontal polarization, and simultaneously record both the copolarized and crosspolarized backscatter. CPRS Doppler, polarimetric and dual-wavelength reflectivity measurements combined with in situ cloud measurements should lead to the development of empirical models that can more accurately classify cloud-particle phase and habit, and make better quantitative estimates of particle size distribution parameters. This dissertation describes the CPRS hardware, and presents colocated 33 GHz and 95 GHz measurements that illustrate the use of dual-frequency measurements to estimate particle size when Mie scattering, is observed in backscatter from rain and ice-phase clouds. Polarimetric measurements are presented as a means of discriminating cloud phase (ice-water) and estimating crystal shape in cirrus clouds. Polarimetric and dual-wavelength observations of insects are also presented with a brief discussion of their impact on the interpretation of precipitation and liquid cloud measurements. In precipitation, Diermendjian's equations for Mie backscatter (1) and the Marshal-Palmer drop-size distribution are used to develop models relating differences in the reflectivity and mean velocity at 33 GHz and 95 GHz to the microphysical parameters of rain. These models are then used to estimate mean droplet size from CPRS measurements of drizzle, which were collected in July, 1993 during the system's first field test in Lincoln, NE. The dissertation also presents cirrus cloud and other measurements collected during the DOE-sponsored Remote Cloud Sensing Intensive Operations Period (RCS-IOP) experiment in April, 1994. Zenith-pointing cirrus measurements show small differences in 33 GHz and 95 GHz reflectivity, as models have predicted (2). Depolarization was also detected in a few cases when ice crystals precipitated from the base of a cloud. On May 29, 1994 CPRS observed a convective storm that produced a cirrus anvil cloud and hail. These storms are one 'engine' producing cirrus clouds and are currently a topic of intensive research by climatologists. Both zenith-pointing and range-height data formats are presented. Measurements of depolarization above the melting/layer are compared to in situ observations of particle size and shape. The RCS-IOP experiment also provided a first opportunity to verify our calibration with aircraft in situ measurements, and to compare our cloud measurements to those collected by other remote sensors. (Abstract shortened by UMI.)

This dissertation classifies trees, snow, and clouds using multiparameter millimeter-wave radar data at 35, 95, and 225 GHz. Classification techniques explored include feedforward multilayer perceptron neural networks trained with standard backpropagation, Gaussian and minimum distance statistical classifiers, and rule-based classifiers. Radar data products, serving as features for classification, are defined, radar and in situ data are presented, scattering phenomenology is discussed, and the effect of data biases are analyzed. A neural network was able to discriminate between white pine trees and other broader-leaved trees with an accuracy of 97% using normalized Mueller matrix data at 225 GHz; wet, dry, melting, and freezing snow could be discriminated 89% of the time using 35, 95, and 225 GHz Mueller matrix data; and metamorphic and fresh snow could be differentiated 98% of the time using either the copolarized complex correlation coefficient or normalized radar cross section at three frequencies. A neural network was also able to discriminate ice clouds from water clouds using vertical and horizontal 95 GHz airborne reflectivity measurements with a success rate of 82% and 86% when viewing the clouds from the side and below respectively. Using 33 and 95 GHz data collected from the ground, a neural net was able to discriminate between ice clouds, liquid clouds, mixed phase clouds, rain, and insects 95% of the time using linear depolarization ratio, velocity, and range. As a precursor to this classification, a rule-based classifier was developed to label training pixels, since in situ data was not available for this particular data set. Attenuation biases in reflectivity were also removed with the aid of the rule-based classifier. A neural network using reflectivity in addition to other features was able to classify pixels correctly 96% of the time.

The Engineering Research Center for Collaborative Adapting Sensing of the Atmosphere (CASA) was established to improve the coverage of the lowest portion of the atmosphere through coordinated scanning of low-power, short-range, networked radars (referred to as Distributed Collaborative Adaptive Sensing (DCAS)). The first DCAS technology demonstration test-bed has been deployed in south-west Oklahoma in early 2006: a network of four, low-power, short-range, dual polarization, Doppler radar units, referred to as IPI (after CASA’s Integrated Project 1). This dissertation is devoted to documenting the IP1 system. Special emphasis is placed on the aspects that enable coordinated radar operation and on other features that provide substantial improvements over existing approaches. In particular, the IP1 radar network can sample the atmosphere with high spatio-temporal resolution and at low altitudes. The dual polarization capabilities and simultaneous multiple radar observations of weather phenomena enable the retrieval of enhanced data products including attenuation corrected reflectivity, dual polarization parameters, and vector wind fields. In addition, the modular radar control, data processing, and communications software architecture allows variations in the network topology, control, and weather information extraction, making the extension of the network easy through the addition of potentially heterogeneous radar nodes.

For over 30 years, radars have examined the structure of the convective boundary layer (CBL). Those studies have consisted either of the three dimensional structure of km-scale features, or of the vertical structure of local, 1 to 100 m-scale features. A new instrument, the Turbulent Eddy Profiler (TEP), images the local, three dimensional character of the CBL with the 10 m-scale resolution of current vertically profiling systems. This thesis presents TEP CBL measurements, including $\tilde C\sbsp{n}{2}$, the local refractive index structure-function parameter, and w, the vertical velocity. Qualitative horizontal and vertical images are shown. The scales of the measured structures are then quantified through calculation of the correlation distance. To examine larger scale features, effective volumes are constructed from TEP time series data through Taylor's hypothesis. Within those volumes, the statistical properties of $\tilde C\sbsp{n}{2}$ and w and calculated. These measurements highlight some of the capabilities of the TEP system, and give a unique picture of the morphology and evolution of $\tilde C\sbsp{n}{2}$ and w in the CBL. Many of the TEP measurements are compared to appropriately scaled large-eddy simulation (LES) predictions. The LES qualitative CBL structure agrees well with the measurements, while the statistical values of $\tilde C\sbsp{n}{2}$ agree well for only some of the measured data. Those $\tilde C\sbsp{n}{2}$ comparisons are the first of their kind, however, and suggest that LES may become a useful tool in CBL propagation studies.

This dissertation presents a unique Compact Millimeter-wave Radar (CMR) developed at the University of Massachusetts Microwave Remote Sensing Laboratory (MIRSL) for measurements of clouds and precipitation. CMR is a 95 GHz solid-state radar with high resolution and Doppler capability which is intended for the eventual use on an unmanned aerospace vehicle (UAV). Details of the design are given here with emphasis on the approaches used to reduce the size and power consumption by almost an order of magnitude compared to previous cloud radars. Ground-based experiments were conducted comparing measurements of clouds and snow made with CMR to those made with the Cloud Profiling Radar System, a full-sized 95 GHz radar with greater sensitivity. The results are used to verify the performance of CMR and to provide an initial external calibration.

This dissertation presents the design and operation of a high-resolution frequency-modulated continuous-wave (FM-CW) radar system to study the structure and dynamics of clear-air turbulence in the atmospheric boundary layer (ABL). This sensitive radar can image the vertical structure of the ABL with both high spatial and temporal resolutions, and provide both qualitative information about the morphology of clear-air structures and quantitative information on the intensity of fluctuations in refractive-index of air. The principles of operation and the hardware and data acquisition characteristics of the radar are described in the dissertation. In October 1999, the radar participated in the Cooperative Atmosphere-Surface Exchange Study (CASES'99) Experiment to characterize the temporal structure and evolution of the boundary-layer features in both convective and stable conditions. The observed structures include clear-air convection, boundary layer evolution, gravity waves, Kelvin-Helmholtz instabilities, stably stratified layers, and clear-air turbulence. Many of the S-band radar images also show high-reflectivity returns from Rayleigh scatterers such as insects. An adaptive median filtering technique based on local statistics has, therefore, been developed to discriminate between Bragg and Rayleigh scattering in clear-air radar observations. The filter is tested on radar observations of clear air convection with comparison to two commonly used image processing techniques. The dissertation also examines the statistical mean of the radar-measured [special characters omitted] for clear-air convection, and compares it with the theoretical predictions. The study also shows that the inversion height, local thickness of the inversion layer, and the height of the elevated atmospheric layers can be estimated from the radar reflectivity measurements. In addition, comparisons to the radiosonde-based height estimates are made. To examine the temporal and spatial structure of [special characters omitted], the dissertation presents two case studies with the measurements of remote (the FM-CW radar and Doppler lidar) and in-situ (research aircraft, kite, and radiosonde) sensors from the stable nighttime boundary layer. It also presents a unique observation of evolution of the convective and nocturnal boundary layers by the S-band radar, and provides description of the observed boundary layer characteristics with the aid of in-situ measurements by the 55m instrumented tower and radiosonde.

This dissertation analyzes the retrieval of water vapor profiles via microwave radiometry; in particular it determines the information content of spectral data and identifies optimal measurement frequencies using an information content technique. The vertical resolution and estimate variance of water vapor profiles derived from the linear inversion of atmospheric data is examined and the effects of measurement noise on these quantities is considered. The Microwave Remote Sensing Laboratory at the University of Massachusetts has developed a unique K-Band Spectral Radiometer (KSR) system that simultaneously monitors eighteen frequencies near the 22.235 GHz water vapor absorption line and is designed to retrieve atmospheric water vapor density profiles by inverting spectral radiance measurements. This system is unique in its measurement speed and breadth. The dissertation discusses calibration techniques, system parameters, and the derivation of a statistical estimation algorithm is that is applied to KSR measurements taken during a field experiment in Lamont, Oklahoma. The resulting water vapor profiles are presented, along with a comparison of in-situ and independent observations.

This dissertation describes the design, fabrication and deployment of the Compact Microwave Radiometer for Humidity profiling (CMR-H). The CMR-H is a new and innovative spectrometer radiometer that is based on monolithic microwave and millimeter-wave integrated circuit (MMIC) technology and is designed for tropospheric water vapor profiling. The CMR-H simultaneously measures microwave emission at four optimally-selected frequency channels near the 22.235 GHz water vapor absorption line, constituting a new set of frequencies for the retrieval of the water vapor profile. State-of-the-art water vapor radiometers either measure at additional channels with redundant information or perform multi-frequency measurements sequentially. The fabrication of the CMR-H demonstrates the capability of MMIC technology to reduce substantially the operational power consumption and size of the RF and IF sections. Those sections comprise much of the mass and volume of current microwave receivers for remote sensing, except in the case of large antennas. The use of the compact box-horn array antenna in the CMR-H demonstrates its capability to reduce the mass and volume of microwave radiometers, while maintaining similar performance to that of commonly-used, bulky horn antennas. Due to its low mass, low volume, low power consumption, fabrication complexity and cost, the CMR-H represents a technological improvement in the design of microwave radiometers for atmospheric water vapor observations. The field test and validation of the CMR-H described in this work focuses on comparisons of measurements during two field experiments from the CMR-H and a state-of-the-art microwave radiometer, which measures only in a volume subtended by the zenith-pointing antenna's beam pattern. In contrast, the CMR-H is designed to perform volumetric scans and to function correctly as a node in a network of radiometers. Mass production of radiometers based on the CMR-H design is expected to enable the implementation of a dense network of radiometers designed to perform measurements of the 3-D water vapor field, with the potential to improve weather forecasting, particularly the location and timing of the initiation of intense convective activity responsible for potentially damaging winds, rain, hail and lightning.

When an optical wave propagates through atmosphere, the wave experiences amplitude, phase, and angle-of-arrival (AOA) fluctuations which are mainly caused by the refractive-index fluctuations of the atmosphere. Thus, the wave fluctuations carry characteristics of the atmosphere. In this dissertation, the AOA fluctuations are studied theoretically and experimentally. For the theoretical part, closed-form solutions of the AOA fluctuations for plane and spherical waves observed by a receiver with a finite aperture were developed. It was assumed that the waves propagate through homogeneous and isotropic media and that the Rytov approximation is valid. The existing closed-form solutions of the AOA fluctuations for the waves are valid only for the cases that the aperture size of the receiver is much larger or much smaller than the Fresnel length. The closed-form solutions developed in this dissertation, however, are valid for all ratios between the aperture diameter and the Fresnel length. The closed-form solutions were compared with the numerical solutions and the accuracy of the closed-form solutions is less than 0.2%. For the experimental part, remote sensing of wind speed transverse to a propagation path using the frequency spectra of the AOA fluctuations was performed with a telescope and a CCD camera. The knee frequency, the intersection of the -2/3 and -8/3 power laws of the spectrum, is a function of wind speed and an effective baseline. If the knee frequency and the effective baseline are known, the transverse wind speed can be retrieved, and if the knee frequency and the transverse wind speed are known, the effective baseline can be estimated. From the measured knee frequency of the spectra of the AOA fluctuations and the aperture size of the telescope as initial guess for the effective baseline, the path-averaged transverse wind speed was retrieved. The effective baseline was calibrated based on wind speed measured by a anemometer. The rms difference between the path-averaged calibrated wind speed retrieved from the frequency spectra of the AOA fluctuations with 30 s of estimation time and the 30 s time-averaged transverse wind speed measured by the anemometer was 11 cm/s -1, while the wind speed varied between 0 and 80cm/s-1.