The dissertation first investigates the single-scattering properties of inhomogeneous ice crystals containing air bubbles. Specifically, a combination of the ray-tracing technique and the Monte Carlo method is used to simulate the scattering of light by randomly oriented large hexagonal ice crystals containing spherical or spheroidal air bubbles. The effect of the air bubbles within ice crystals is to smooth the phase functions, diminish the 22° and 46° halo peaks, and reduce the backscatter in comparison with the case of bubble-free ice crystals. Cloud reflectance look-up tables were generated at the wavelengths of 0.65 μm and 2.13 μm to examine the impact of accounting for air bubbles in ice crystal morphology on the retrieval of ice cloud optical thickness and effective particle size. To investigate the effect of the representation of aggregates on electromagnetic scattering calculations, an algorithm is developed to efficiently specify the geometries of aggregates and to compute some of their geometric parameters such as the projected area. Based on in situ observations, aggregates are defined as clusters of hexagonal plates with a chain-like overall shape. An aggregate model is developed with 10 ensemble members, each consisting of between 4-12 hexagonal plates. The scattering properties of an individual aggregate ice particle are computed using the discrete dipole approximation or an Improved Geometric Optics Method, depending upon the size parameter. The aggregate model provides an accurate and computationally efficient way to represent all aggregates occurring within ice clouds. We developed an algorithm to determine an appropriate ice cloud model for application to satellite-based retrieval of ice cloud properties. Collocated Moderate Resolution Imaging Spectroradiometer and Multi-angle Imaging SpectroRadiometer (MISR) data are used to retrieve the optical thicknesses of ice clouds as a function of scattering angle in the nine MISR viewing directions. The difference between cloud optical thickness and its averaged value over the nine viewing angles can be used to validate the ice cloud models. Using the data obtained on 2 July 2009, an appropriate ice cloud model is determined. With the presence of all the uncertainties in the current operational satellite-based retrievals of ice cloud properties, this ice cloud model has excellent performance in terms of consistency in cloud property retrievals with the nine MISR viewing angles.

During the summer of 2002, a modified tandem differential mobility analyzer (TDMA) was used to examine the size-resolved hydration state of the ambient aerosol in Southeast Texas. Although there were slight variations in the measured properties over the course of the study, the deliquescent particles observed were almost always present as metastable aqueous solutions. A relative humidity (RH) scanning TDMA system was used to measure the deliquescence/crystallization properties of ambient aerosol populations in the same region. During August, sampling was conducted at a rural site in College Station, and in September at an urban site near the Houston ship channel. Measurements from both sites indicate cyclical changes in the composition of the soluble fraction of the aerosol, which are not strongly linked to the local aerosol source. The observations show that as temperature increases and RH decreases, the hysteresis loop describing the RH-dependence of aerosol hygroscopic growth collapses. It is proposed that this collapse is due to a decrease in the ammonium to sulfate ratio in the aerosol particles, which coincides with increasing temperature and decreasing RH. This cyclical change in aerosol acidity may influence secondary organic aerosol (SOA) production and may exaggerate the impact of the aerosol on human health. The compositional changes also result in a daily cycle in crystallization RH that is in phase with that of the ambient RH, which reduces the probability that hygroscopic particles will crystallize in the afternoon when the ambient RH is a minimum. During June and July of 2004 airborne measurements of size-resolved aerosol hygroscopic properties were made near Monterey, California. These were used to examine the change in soluble mass after the aerosol had been processed by cloud. The calculated change in soluble mass after cloud-processing ranged from 0.66 g m-3 to 1.40 g m-3. Model calculations showed these values to be within the theoretical bounds for the aerosols measured. Mass light-scattering efficiencies were calculated from both an averaged aerosol size distribution and from distributions modified to reflect the effects of cloud. These calculations show that the increase in mass light-scattering efficiency should be between 6% and 14%.

In this study, we use thermal infrared channels to retrieve the optical thickness and effective particle radius of ice clouds. A physical model is used in conjunction with Atmospheric Infrared Sounder (AIRS) temperature and water vapor profiles to simulate the top-of-atmosphere (TOA) brightness temperatures (BTs) observed by the Moderate Resolution Imaging Spectroradiometer (MODIS) for channels located at 8.5, 11.0, and 12.0 µm (1176, 909, and 833 cm-1). The model is initially validated by comparing simulated clear-sky BTs to MODIS-observed clear-sky BTs. We also investigate the effect of introducing a +3 K bias in the temperature profile, a +3 K bias in the surface temperature, and a +20% bias in the water vapor profile in order to test the sensitivity of the model to these inputs. For clear-sky cases, the simulated TOA BTs agree with MODIS to within 2-3 K. The model is then extended to simulate thermal infrared BTs for cloudy skies, and we infer the optical thickness and effective radius of ice clouds by matching MODIS-observed BTs to calculations. The optical thickness retrieval is reasonably consistent with the MODIS Collection 5 operational retrieval for optically thin clouds but tends to retrieve smaller particle sizes than MODIS.

A field-deployable instrument has been developed that isolates from an ambient aerosol population only those particles that have critical supersaturations, Sc, within a narrow, user-specified, range. This Differential Critical Supersaturation Separator (DScS) is designed to supply one or more particle size and/or composition analyzers to permit the direct examination of the factors that influence the activation properties of ambient aerosols. The DScS consists of two coupled parallel plate continuous flow thermal gradient diffusion cloud chambers housed within a single enclosure. Descriptions of instrument operation, construction and calibration data collected, when pure ammonium sulfate aerosols were injected into the DScS for operation at 0.15%< Sc<0.175%, 0.30%< Sc<0.35%, and 0.60% < Sc<0.70%, are included. Following instrument development, the DScS was deployed during March 2006 for the Megacities Impact on Regional And Global Environment (MIRAGE) field campaign in Mexico City, Mexico. Throughout the MIRAGE field campaign a Differential Mobility Analyzer/Tandem Differential Mobility Analyzer (DMA/TDMA) system measured aerosol size distributions and size-resolved hygroscopicity of DScS separated aerosol. The dry diameter (Dp*) of particles sampled in the TDMA system as well as the known Sc prescribed in the DScS were combined in a modified version of Köhler Theory to make predictions of particle hygroscopicity. These predictions frequently overestimated the measurements. Further analysis of DScS separated aerosols compares the known particle Sc to a predicted particle Sc, providing insight into particle activation efficiency. Overall, the sampled aerosol exhibited properties that indicate they were more efficient at activation than Köhler Theory would predict.

The Houston LDAR II network is an array of twelve VHF time-of-arrival (TOA) sensors operated by Texas A&M University. The goals of the network are to conduct indepth studies of thunderstorm electrification and provide timely lightning threats to the Houston area. Before analyses are conducted using data from the Houston LDAR network, it is necessary to understand the LDAR networkâ s performance and limitations, such as the LDAR source detection efficiency, network range, and location accuracy. Initial results from the 31 October 2005 Mesoscale Convective System (MCS) timing error analysis revealed an RMS timing error for the Houston LDAR network of 90 ns for 6 sensor solutions. This gives a three-dimensional location accuracy of 1 km at a distance of 150 km and 100 m over the center of the network. Reanalysis with updated sensor positions decreased the RMS timing error to 75 ns. This decrease in RMS timing error increased the median three-dimensional location accuracy by ~100 m at a 100 km range. The network has been operated at both 70 MHz and 40 MHz. Model results of detection efficiency suggest that the change to 40 MHz yields an increase of 9 - 10 dB in network sensitivity. Analysis of VHF source power distributions shows a similar shift from that expected from the model. These results show that the 40 MHz LDAR network detects ~3.3 times more VHF sources than the 70 MHz network. The analysis of the usable network range for research purposes is currently set by rough guidelines of location accuracy and detection efficiency. For location accuracy, a 1 km limit allows storm analysis out to a range of more than 150 km. For the detection efficiency analysis, results based on source power distributions suggest a gradual fall off with range. Examining the change in VHF source density by range reveals different results. VHF source density remained fairly constant out to a range of 100 km at which point a significant fall off was observed. Based on these results the usable network range for the Houston network is 100 km.

Thesis (Ph.D)--Earth and Atmospheric Sciences, Georgia Institute of Technology, 2009. / Committee Chair: Athanasios Nenes; Committee Member: Andrew G. Stack; Committee Member: Irina N. Sokolik; Committee Member: Judith A. Curry; Committee Member: Mike Bergin; Committee Member: Rodney J. Weber. Part of the SMARTech Electronic Thesis and Dissertation Collection.

In addition to the wireless network providers’ need for traffic classification, the need is more and more common in the Cloud Computing environment. A data center hosting Cloud Computing services needs to apply priority policies and Service Level Agreement (SLA) rules at the edge of its network. Overwhelming requirements about user privacy protection and the trend of IPv6 adoption will contribute to the significant growth of encrypted Cloud Computing traffic. This report presents experiments focusing on application of data mining based Internet traffic classification methods to classify encrypted Cloud Computing service traffic. By combining TCP session level attributes, client and host connection patterns and Cloud Computing service Message Exchange Patterns (MEP), the best method identified in this report yields 89% overall accuracy. / text

published_or_final_version / Computer Science / Master / Master of Philosophy

Cloud computing has rapidly emerged as a new computation paradigm, providing agile and scalable resource access in a utility-like fashion. Processing of massive amounts of data has been a primary usage of the clouds in practice. While many efforts have been devoted to designing the computation models (e.g., MapReduce), one important issue has been largely neglected in this respect: how do we efficiently move the data, practically generated from different geographical locations over time, into a cloud for effective processing? The usual approach of shipping data using hard disks lacks flexibility and security. As the first dedicated effort, this paper tackles this massive, dynamic data migration issue. Targeting a cloud encompassing disparate data centers of different resource charges, we model the cost-minimizing data migration problem, and propose efficient offline and online algorithms, which optimize the routes of data into the cloud and the choice of the data center to aggregate the data for processing, at any give time. Three online algorithms are proposed to practically guide data migration over time. With no need of any future information on the data generation pattern, an online lazy migration (OLM) algorithm achieves a competitive ratio as low as 2:55 under typical system settings, and a work function algorithm (WFA) has a linear 2K-1 (K is the number of data centers) competitive ratio. The rest one randomized fixed horizon control algorithm (RFHC) achieves 1+ 1/(l+1 ) κ/λ competitive ratio in theory with a lookahead window of l into the future, where κ and λ are protocol parameters. We conduct extensive experiments to evaluate our online algorithms, using real-world meteorological data generation traces, under realistic cloud settings. Comparisons among online and offline algorithms show a close-to-offline-optimum performance and demonstrate the effectiveness of our online algorithms in practice. / published_or_final_version / Computer Science / Master / Master of Philosophy

Client devices operating at the edges on the Internet, in homes, cars, offices, and elsewhere, are highly heterogeneous in terms of their hardware configurations, form factors, and capabilities, ranging from small sensors to wearable and mobile devices, to the stationary ones like smart TVs and desktop machines. With recent and future advances in wireless networking allowing all such devices to interact with each other and with the cloud, it becomes possible to combine and augment capabilities of individual devices via services running at the edge - in edge clouds - and/or via services running in remote datacenters. The virtual platform approach to combining and enhancing such devices developed in this research makes possible the creation of innovative end user services, using low-latency communications with nearby devices to create for each end user exactly the platform needed for current tasks, guided by permissions and policies controlled by remote, cloud-resident social network services (SNS). To end users, virtual platforms operate beyond the limitations of individual devices, as natural extensions of those devices that offer improved functionality and performance, with ease-of-use provided by cloud-level global context and knowledge.