Calendar

Dissertation Defense

Candidate: Mohammed Sikder
Title: A Regional Crops Forecasting Model Integrating Satellite Remote Remote Sensing with the Localized EPIC Model

Abstract:

Airborne and satellite remote sensing data adds tremendous value in assessing crops phenology, green biomass, and yield forecasting. Also, satellite-based real/near-real time data is used for monitoring growth rates, water and nutrient status, and crop responses to biotic and abiotic stress conditions which could assist with efficient crop management decisions. The objectives of the study were to develop a regional crop yield and area estimation forecasting model for winter wheat in the Texas Panhandle integrating satellite remote sensing data with the localized EPIC crop forecasting model. The Normalized Difference Vegetation Index (NDVI) pixel-level precision data from the Landsat 7 (ETM+) sensor  spanning 5 years (2000, 2005, 2012, 2013, and 2014) during the period 2000-2014 for five counties of the Texas Panhandle region (~14,948 sq. km.) were used for this research and model development efforts. As part of the study, a sophisticated Java Application Programming Interface (API)-based set of toolkits called the Hi-Speed Interactive Landsat Image Processing (HILIP) was developed for processing a massive amount of satellite data as part of an Integrated EPIC model. The software is capable of doing complex image processing including supervised and Landsat-NDVI-based classification with almost 80-90% overall accuracy. This tool is also loaded with the capabilities for pixel-level geophysical positioning and area computation. In this study, an integrated Yearly Winter Wheat Yield prediction model has been developed. In this effort, USDA-NASS reports, adjusted with the EPIC simulated yearly winter wheat yield (bushels/acre) was integrated with the county total harvested winter wheat area (acres) calculated by the Landsat 7 (ETM+) using an NDVI-based model giving the total winter wheat yield in bushels for a specific county in a specified year. Additionally, a statistical winter wheat yield forecasting model was also developed. In this model the county-specific average yearly yield was calculated by taking past 15 years of yield data for that county, giving the yield in bushels/acre, and then multiplying by the Landsat 7 (ETM+) NDVI model-based total harvested winter wheat area for that county, resulting in the total winter wheat yield in bushels for that county for the specified year. The results suggest that the overall accuracy of the Integrated EPIC-NDVI model is within the 80-90% range and the accuracy of the NDVI-NASS statistics-based model is within 70-80%.

Director:

Dr. Ruixin Yang

Committee:

Dr. Daniel B. Carr, Dr. J Qu, Dr. Fernando Camelli

Notes: The thesis is on reserve in the Johnson Center Library, Fairfax Campus. All members of the George Mason University community are invited to attend.

Dissertation Defense

Candidate: Phillip Hess
Title: Understanding the Evolution and Propagation of Coronal Mass Ejections and Associated Plasma Sheaths in Interplanetary Space

Abstract:

A Coronal Mass Ejection (CME) is an eruption of magnetized plasma from the Corona of the Sun. Understanding the physical process of CMEs is a fundamental challenge in solar physics, and is also of increasing importance for our technological society. CMEs are known the main driver of space weather that has adverse effects on satellites, power grids, communication and navigation systems and astronauts. Understanding and predicting CMEs are still in the early stage. In this dissertation, much improved observational methods and advanced theoretical analysis are used to study CMEs.

Unlike many studies in the past that treat CMEs as a single object, this study divides a CME into two separate components: the ejecta from the corona and the sheath region that is the ambient plasma compressed by the shock/wave running ahead of the ejecta; both structures are geo-effective but evolve differently. Stereoscopic observations from multiple spacecraft, including STEREO and SOHO, are combined to provide a three-dimensional geometric reconstruction of the structures studied. True distances and true vector velocities of CMEs are accurately determined, free of projection effect, and with continuous tracking from the low corona to 1 AU.

To understand the kinematic evolution of CMEs, an advanced drag-based model (DBM) is proposed, with several key improvements to the original DMB model. First, the new model allows the drag parameter to vary with distance; the variation is constrained by the necessary conservation of physical parameters. Second, the deviation of CME-nose from the Sun-Earth-line is taken into account. Third, the geometric correction of the shape of the ejecta front is considered, based on the assumption that the true front is a flattened croissant-shaped flux rope front.

These improvements of the DBM model provide a framework for using measurement data to make accurate prediction of the arrival times of CME ejecta and sheaths. Using a set of seven events to test the model, it is found that the evolution of the ejecta front can be accurately predicted, with a slightly poorer performance on the sheath front. To improve the sheath prediction, the standoff-distance between the ejecta and the sheath front is used to model the evolution. The predicted arrivals of both the sheath and ejecta fronts at Earth are determined to within an average 3.5 hours and 1.5 hours of observed arrivals, respectively. These prediction errors show a significant improvement over predictions made by other researchers. The results of this dissertation study demonstrate that accurate space weather prediction is possible, and also reveals what observations are needed in the future for realistic operational space weather prediction.

Director or Committee Chair:

Dr. Jie Zhang

Committee:

Dr. Dusan Odstrcil
Dr. Arthur Poland
Dr. Robert Weigel
Dr. Chi Yang

Notes: The thesis is on reserve in the Johnson Center Library, Fairfax Campus. All members of the George Mason University community are invited to attend.

Dissertation Defense

Candidate: David Hamilton
Title: Machine-Readable Knowledge Management of Neuron Properties

Abstract:

The advancement of neuroscience, perhaps one of the most information rich disciplines of all the life sciences, benefits greatly from neuroinformatic techniques to manage the vast amounts of data generated by the research community to promote novel insights and integrated understanding. Since Cajal, the neuron remains a fundamental unit of the nervous system, yet even with the explosion of information technology, we still have few comprehensive or systematic strategies for aggregating cell-level knowledge.

Widely spread naming inconsistencies in neuroscience pose a vexing obstacle to effective communication within and across areas of expertise. This problem is particularly acute when identifying neuron types and their properties. Hippocampome.org is a web-accessible neuroinformatics resource that organizes existing data about essential properties of all known neuron types in the rodent hippocampal formation. Hippocampome.org links evidence supporting the assignment of a property to a type with direct pointers to quotes and figures.

Mining this knowledge from peer-reviewed reports and Allen Brain Atlas (ABA) Mouse Brain (MB) in situ hybridization (ISH) data, reveals the troubling extent of terminological ambiguity and undefined terms. Examples span simple cases of using multiple synonyms and acronyms for the same molecular biomarkers (or other property) to more complex cases of neuronal naming. New publications often use different terms without mapping them to previous terms. As a result, neurons of the same type are assigned disparate names, while neurons of different types are bestowed the same name.

Furthermore, non-unique properties are frequently used as names, and several neuron types are not named at all. In order to alleviate this nomenclature confusion regarding hippocampal neuron types and properties, we introduce a new functionality of Hippocampome.org: a fully searchable, curated catalog of human and machine-readable definitions, each linked to the corresponding neuron and properties terms. Furthermore, we extend our robust approach to providing each neuron type with an informative name and unique identifier by mapping all encountered synonyms and homonyms.

In Hippocampome.org, given the biochemical profile is relatively sparse even for the best studied neuron types, we choose to extend these characterizations by leveraging the massive mouse brain gene expression analysis conducted by the Allen Institute. The Allen Brain Atlas (ABA) provides a wealth of data that when appropriately interpreted can be leveraged to substantially augment the biomarker knowledge in Hippocampome.org. In this study, we restrict the investigation to the principal cell layers of dentate gyrus (DG), CA3, CA2, and CA1, where the vast majority of the neurons are glutamatergic projecting neurons. Thus, ABA Mouse Brain (MB) in situ hybridization (ISH) data for those layers can justifiably be linked to the respective principal neuron types (i.e. Granule in DG and Pyramidal in CA3, CA2, and CA1). We filtered the whole-genome expression dataset to maximize consistency with current Hippocampome.org biomarker content. The resulting additional set of expressed/not-expressed genes expands the ~1K Hippocampome.org biomarker pieces of knowledge (PoK) by ~5K, yielding a considerably more complete genetic characterization of principal neuron types in the mouse hippocampus.

Committee Chair:

Dr. Giorgio A. Ascoli

Committee:

Dr. James L. Olds
Dr. J. Robert Cressman
Dr. Kenneth P. Smith

Notes: The thesis is on reserve in the Johnson Center Library, Fairfax Campus. All members of the George Mason University community are invited to attend.

Dissertation Defense

Candidate: Kelly Hamilton
Title: Basic Helix Loop Helix Enhancer 40 in Neuronal Excitability and Synaptic Plasticity

Abstract:

This dissertation describes the role of the Basic Helix Loop Helix Enhancer 40 (Bhlhe40) transcription factor in the adult murine brain at the molecular, cellular, network, and behavioral levels. Studies for this dissertation were performed on a congenic Bhlhe40 gene knock out mouse model (Bhlhe40 KO). The inspiration for this research project was based on prior findings in mice that were genetically null for the Bhlhe40 gene on a mixed genetic background (Jiang et al., 2008, J. Neurosci). Mixed background Bhlhe40 KO mice had enhanced seizure activity when injected intraperitoneally with the convulsant, Kainic Acid (KA). Changes in neuronal gene expression occur as a result of seizure activity, particularly in the hippocampus and in the gene encoding brain-derived neurotropohic factor (BDNF). In the hippocampus, BDNF levels are increased following seizure activity and are tho ught to lower the threshold for subsequent seizures, implicating BDNF in a positive feedback loop in epileptogenesis. Specifically, mixed background Bhlhe40 KO mice had elevated basal levels of BDNF-4 transcripts, which are normally expressed in an activity-dependent manner. The central hypothesis of this research was that congenic Bhlhe40 KO mice would have enhanced responses to KA-induced seizures due to excessive levels of basal BDNF. It was further thought that there would be coincident increases in neuronal activity in hippocampal slices and increased expression of genes modulating neuronal excitability.

The first objective of this research was to elucidate changes in gene expression occurring in the hippocampus of congenic Bhlhe40 KO mice. A whole genome expression array was utilized to capture an unbiased profile of hippocampal mRNA levels from Bhlhe40 KO mice compared to wild type (WT) mice. Gene expression array findings were independently validated by quantitative gene specific mRNA and protein assays. I found that mRNA and protein levels for Insulin Degrading Enzyme (IDE) were two-fold decreased in congenic Bhlhe40 KO hippocampi. Unlike in the mixed background Bhlhe40 KO mice, congenic Bhlhe40 KO hippocampi did not have elevated levels of BDNF mRNA or protein levels.

At the cellular level, I sought to determine the role of Bhlhe40 KO in neuronal excitability. To test this, Bhlhe40 KO hippocampal CA1 neurons, cells that express Bhlhe40 in WT mice, were measured for excitatory and inhibitory electrophysiological properties, and were determined to have enhanced excitation and reduced inhibition, indicating a hyperexcitable state. At the network level, I tested Bhlhe40 KO hippocampal slices for changes in synaptic plasticity and found a decrease in both Long Term Potentiation (LTP) and Long Term Depression (LTD), indicating an overall reduction in long term synaptic plasticity.

At the behavioral level, I tested seizure severity in Bhlhe40 KO mice by intrahippocampal KA-injection, but despite the increase in excitability on the cellular level I found no significant difference in seizure response between Bhlhe40 KO mice and WT controls. In addition, anxiety and learning and memory performance was determined in untreated congenic Bhlhe40 KO and WT mice. Despite the reduction in synaptic plasticity, no changes in spatial learning and memory were observed on the Morris Water Maze, however, there was an increase in anxiety behavior seen on the Open Field.

An interesting finding from this work was the effect of genetic background, namely in regards to seizure susceptibility and BDNF expression. Inter-strain differences can be explained at the genomic level by variation in promoter and other regulatory sequences in the genome. Importantly, I propose here that changes in IDE protein levels may be driving changes in basal excitability and reduced synaptic responses to stimulation, as well as the anxiety phenotype. Insulin levels were investigated and found to be non-significantly changed in Bhlhe40 KO hippocampus, however, IDE also degrades Insulin-like Growth Factor -1 & -2, and Amyloid beta. Further investigation into these other IDE substrates may elucidate the link between Bhlhe40-mediated IDE regulation, anxiety, neuronal excitability and synaptic plasticity.

Dissertation Director:

Dr. Robert H. Lipsky

Committee:

Dr. Mark P. Mattson
Dr. Ann Marini
Dr. Daniel N. Cox
Dr. Nadine Kabbani

Notes: The thesis is on reserve in the Johnson Center Library, Fairfax Campus. All members of the George Mason University community are invited to attend.

Thesis Defense

Candidate: Stephanie Barksdale
Title: Novel Antimicrobial Peptides in Alligator and Crocodile

Abstract:

Novel antibiotics are needed to fight the rising tide of drug resistance in pathogenic bacteria. One possible source is cationic antimicrobial peptides (AMPs), small proteins produced by the innate immune system. AMPs have a range of mechanisms including direct antibacterial action and immunomodulatory effects.

Crocodilians are part of an ancient clade, the Archosaurs, and are more closely related to birds and dinosaurs than other living reptiles. Very little is known about the innate immune systems of crocodilians, but research has found that the serum of these species have antimicrobial activity beyond that of human serum. This activity is thought to be partly due to AMPs, though only a handful of crocodilian AMPs have been described.

In this thesis, four novel AMPs from members of the order Crocodilia are investigated. A hepcidin from Crocodylus siamensis, an iron-regulating peptide with 4 intramolecular disulfide bonds, is found to have weak activity against Pseudomonas aeruginosa, Escherichia coli, and Staphylococcus aureus. Two fragments of an apolipoprotein found in the blood of Alligator mississippiensis were found to have strong activity against a range of Gram negative and Gram positive bacteria, including multi-drug resistant bacteria. These fragments were found to be alpha-helical and to depolarize the bacterial membrane. A cathelicidin from A. mississippiensis is strongly active against P. aeruginosa and multi-drug resistant Acinetobacter baumannii and forms pores in the bacterial membrane.

These analyses give us greater understanding of the crocodilian innate immune system. In addition, these AMPs could be used as a basis for new antimicrobials.

Thesis Director:

Dr. Monique van Hoek

Committee:

Dr. Serguei Popov
Dr. Barney Bishop

Notes: The thesis is on reserve in the Gateway Library, Science and Technology Campus. All members of the George Mason University community are invited to attend.

Dissertation Defense

Candidate: Lisa Frances LaCivita
Title: Amphibian Monitoring for Ecosystem Services, Citizen Engagement and Public Policy

Director or Committee Chair:

Dr. Thomas E. Lovejoy

Committee:

Dr. Lee Talbot
Dr. R. Christian Jones
Dr. K. Bruce Jones

Notes: The thesis is on reserve in the Johnson Center Library, Fairfax Campus. All members of the George Mason University community are invited to attend.

Dissertation Defense

Candidate: Manzhu Yu
Title: Spatiotemporal Methodologies and Analytics for Extreme Weather Study-Using Dust Storm Even as an Example

Abstract:

Dust storm represents a serious hazard to health, property, and the environment in arid and semi-arid areas. To mitigate the hazardous impact of dust storms, it is crucial to detect an upcoming dust event and predict its evolution to inform the early warning and decision-making process. In this dissertation, research is presented into spatiotemporal methods for addressing the problems and challenges, such as dust model uncertainty, challenges of automatically identifying dust features, tracking the evolution pattern of dust events, and challenges in spatiotemporal data modeling.
This dissertation makes innovative contribution for the following reasons: 1) integrating spatiotemporal thinking into the improvement of dust modeling, transforming the spatiotemporal variations of input from static to dynamic; 2) extending feature identification and tracking of dust features into a higher dimensionality, from 2D/2.5D to 3D/4D; 3) proposing a spatiotemporal data framework to better capture the evolution and transport paths of natural phenomena; and 4) integrating GIScience with geographic studies, computational science and geoscience together for a multidisciplinary study. The result of this dissertation can be helpful to provide insights on improving dust forecasting and interpreting simulation data for early warning of dust events, and ultimately provide information for real applications that sustain human lives and resources.

Director or Committee Chair:

Chaowei Yang

Committee:

George Taylor
Ruixin Yang
Dieter Pfoser
Songqing Chen

Notes: The thesis is on reserve in the Johnson Center Library, Fairfax Campus. All members of the George Mason University community are invited to attend.

Dissertation Defense

Candidate: Daniel Ray Sponseller
Title: Molecular Dynamics Study of Polymers and Atomic Clusters

Abstract:

This dissertation contains investigations based on Molecular Dynamics (MD) of a variety of systems, from small atomic clusters to polymers in solution and in their condensed phases. The overall research is divided in three parts. First, I tested a new thermostat in the literature on the thermal equilibration of a small cluster of Lennard-Jones (LJ) atoms. The proposed thermostat is a Hamiltonian thermostat based on a logarithmic oscillator with the outstanding property that the mean value of its kinetic energy is constant independent of the mass and energy. I inspected several weak-coupling interaction models between the LJ cluster and the logarithmic oscillator in 3D. In all cases I show that this coupling gives rise to a a kinetic motion of the cluster center of mass without transferring kinetic energy to the interatomic vibrations. This is a failure of the published thermostat because the temperature of the cluster is mainly due to vibrations in small atomic clusters This logarithmic oscillator cannot be used to thermostat any atomic or molecular system, small or large.

Director or Committee Chair:

Estela Blaisten-Barojas, Dissertation Director

Committee:

Howard Sheng
James Glasbenner
Dmitri Klimov

Notes: A copy of Dan’s dissertation is available for examination from Karen Underwood, Department of Computational and Data Sciences, 373 Research Hall. The dissertation is available to read only within the Department and cannot be taken out of the Department or copied.

Dissertation Defense

Candidate: Muhammad N. Baqui
Title: Automated Monitoring of High Density Crowd Events

Abstract:

Pedestrian traffic is an important subject of surveillance to ensure public safety and traffic management, which may benefit from intelligent and continuous analysis of Pedestrian traffic  videos. State-of- the-art methods for intelligent pedestrian traffic surveillance have a number of limitations in automating and computing useful pedestrian traffic information from closed circuit television (CCTV) images. This work aims to automate and augment the traditional pedestrian traffic surveillance system by introducing four components in a novel integrated framework. A fast and efficient particle image velocimetry (PIV) technique is proposed to yield pedestrian velocities for timely management of pedestrian traffic. A machine learning-based regression model, boosted Ferns, is used to improve pedestrian count estimation: an essential metric for high-density pedestrian traffic analysis. A camera perspective model is proposed to improve the speed and position estimate of pedestrian traffic incorporating camera’s intrinsic and extrinsic parameters. All these functional improvements are integrated in a seamless framework to predict future pedestrian traffic distribution, which is a crucial piece   of information for pedestrian traffic management. The proposed framework is computationally efficient, suitable for multiple camera feeds with high-density pedestrian traffic videos, and capable of rapidly analyzing and predicting flows of thousands of pedestrians. The proposed framework is one of the first steps towards fully integrated CCTV-based automated pedestrian traffic management system.

Director or Committee Chair:

Rainald Löhner

Committee:

Juan Cebral, Chi Yang, Jan Allbeck

Notes: A copy of Muhammad’s dissertation is available for examination from Karen Underwood, Department of Computational and Data Sciences, 373 Research Hall. The dissertation is available to read only within the Department and cannot be taken out of the Department or copied.

The thesis is on reserve in the Johnson Center Library, Fairfax Campus. All members of the George Mason University community are invited to attend.

Dissertation Defense

Candidate: Brian Corgiat
Title: Molecular mechanisms underlying visual recognition memory in rhesus monkey

Abstract:

Visual recognition memory is the ability to identify previously encountered objects as familiar and enables us to recognize our family and friends and to navigate from one place to another. Visual recognition memory is critically dependent upon the perirhinal cortex (PRh). More specifically, visual memory formation requires cholinergic activation of the m1 muscarinic acetylcholine receptor (m1AChR) in PRh and is also characterized by enhanced multiunit activity in the upper middle and deep PRh layers. However, the M1 muscarinic-dependent intracellular signaling pathways underlying the critical synaptic changes induced during visual memory formation remain unknown.

The first research aim of this dissertation is to develop a proteomic approach to study the intracellular signaling pathways that underlie visual recognition memory in a cortical layer- and subregion- specific manner in nonhuman primate (NHP). Protein phosphorylation is highly dynamic, and phosphoproteins must be stabilized quickly to preserve their in vivo phosphorylation state. Common fixation techniques, such as transcardial perfusion with formalin/paraformaldehyde, are inadequate at preserving in vivo phosphoprotein levels. Additionally, many standard phosphoprotein quantification methods suffer from limited sensitivity (e.g., immunohistochemistry) or require substantial amounts of tissue per assay (e.g., Western blot), rendering them incompatible for studying subregion- and layer-specific signaling in NHP brain regions. A workflow was established able to measure cortical layer-specific phosphoprotein signaling pathways in NHP using laser-capture microdissection and reverse phase protein microarrays. Using this workflow, the impact of two parameters likely to be encountered in future behavioral experiments were assessed: the cold ischemia time post euthanasia and the choice of anesthetic. Further, baseline phosphoprotein signaling in critical brain regions were measured. Ultimately, this first aim established a technical framework for studies of protein signaling networks that underlie cognitive processes in nonhuman primates.

The second research aim was to assess protein signaling activation after visual memory formation in PRh. Monkeys were trained on a delayed nonmatching-to-sample (DNMS) task, a test of recognition memory, to drive cholinergic activation of the m1AChR and activation of m1AChR-depedent protein pathways in PRh. After extensive training on the DNMS task, monkeys were euthanized, and the tissue collected and processed for a cortical layer-specific analysis of m1AChR-dependent protein pathway activation. Initial clustering analyses revealed no dramatic change in m1AChR-dependent pathway activation after memory training, thus a multi-level model will be developed to parse apart m1AChR-dependent protein pathway activation after recognition memory training.

The third and final aim of this dissertation was to assess the impact of an m1AChR-related potassium channel, the Kv7 KCNQ channel, and an m1AChR-dependent signaling pathway, the mechanistic target of rapamycin (mTOR) pathway, using a standard pharmacological approach. KCNQ channel openers and closers and an mTOR pathway inhibitor were directly microinfused into PRh in monkeys performing one-trial visual recognition. Neither the KCNQ channel manipulations nor the mTOR inhibition impaired monkeys performance on one-trial visual recognition.
In summary, the work in this dissertation developed a new approach to study cortical layer-specific protein pathway activation in monkey and this approach was used to assess m1AChR-depedent signaling after visual memory formation. Additionally, other m1AChR-related mechanisms, such as modulation of KCNQ channels, were found to have limited impact on visual recognition.

Director or Committee Chair:

Dr. Lance Liotta

Committee:

Dr. Claudius Mueller
Dr. James L. Olds
Dr. Kim Blackwell
Dr. Janita Turchi

Notes: The thesis is on reserve in the Johnson Center Library, Fairfax Campus. All members of the George Mason University community are invited to attend.