Dissertations / Theses on the topic 'Master of Geographic Information Science'

The explosion of web-based GIS technologies and the opening up of mapping technologies to common citizens in the past decade have resulted in a whole range of VGI communities like OpenStreetMap, Ushahidi and Wikimapia, that are used to assist emergency management operations on a large scale. However, most crowd sourced systems currently being used for disaster recovery have multiple obstacles like accessibility, ease of use, dependency on social media and requirement of special skill sets on the part of the public participants that serve as limitations to the fulfillment of the democratization potential of VGI.

Hence an improved Android mobile application was developed which is much more accessible, usable, reliable without any dependency on social media like Facebook to collect and transmit data, thus not only ensuring participation equality but also universal accessibility to quality and timely geographic information during emergency situations.

Mobile phones have become so popular in navigation. Empirical studies, however, have implied several pitfalls of using these mobile systems. First of all, the small size of the mobile screen fragmentizes the map information so that users have to interact with the display frequently for fear of disorientation. In addition, generated navigation guides with continuous displays of routing information relieves users’ cognitive load, but the excessive reliance on the displayed spatial information keeps the users being mindless of the environment that impacts their acquisition of spatial knowledge. Later, landmarks as important referents were suggested to help users integrate the current surroundings with further decision points to support their sense of direction and cognitive mapping. But the small display limited users to access the landmarks located out of the view. To address this issue, designs attempt to display distant landmarks at the edge of screen as a way to overcome the aforementioned limits. In order to enhance spatial orientation while using mobile devices, this study introduced an improved design that not only display the direction but also the distance concept of distant landmarks by changing in size as an indicator of the distance from a user’s location. Built on this, two kinds of mechanism were designed to present the distance concept by icons of different sizes: one is based on ratio scale that icons change in size continuously based on an established ratio to the actual distance between the user and the distant location. The other mechanism is based on ordinal scale which assigns one of three different sizes to a certain range of distance implying near, middle, and far. A formal user study was carried out to compare efficiency of these two mechanisms in four types of distance comparison tasks. Results show that ordinal icons are more effective than ratio icons in visualizing relative distances between two distant landmarks. But for both mechanisms, users have challenges distinguishing distant landmarks from local landmarks when displayed on screen simultaneously. A further step is to explore some other feasible options of representing distance.

Federal, state and local officials are planning for a (M) 7.8 San Andreas Earthquake Scenario in the Southern California Catastrophic Earthquake Response Plan that would require initial emergency food and water resources to support from 2.5 million to 3.5 million people over an eight-county region in Southern California. However, a model that identifies locations of affected populations—with consideration for social vulnerability, estimates of their emergency logistical resource requirements, and their resource requirements over time—has yet to be developed for the emergency response plan.

The aim of this study was to develop a modeling methodology for emergency logistical resource requirements of affected populations in the (M) 7.8 San Andreas Earthquake Scenario in Southern California. These initial resource requirements, defined at three-days post-event and predicted through a probabilistic risk model, were then used to develop a relative risk ratio and to estimate resources requirements over time. The model results predict an “at-risk” population of 3,352,995 in the eight-county study region. In Los Angeles County, the model predicts an “at- risk” population of 1,421,415 with initial requirements for 2,842,830 meals and 4,264,245 liters of water. The model also indicates that communities such as Baldwin Park, Lancaster-Palmdale and South Los Angeles will have long-term resource requirements.

Through the development of this modeling methodology and its applications, the planning capability of the Southern California Catastrophic Earthquake Response Plan is enhanced and provides a more effective baseline for emergency managers to target emergency logistical resources to communities with the greatest need. The model can be calibrated, validated, generalized, and applied in other earthquake or multi-hazard scenarios through subsequent research.

Electoral districts drawn by independent commissions are seen by political reformers to be preferable to those drawn by state legislatures. The overtly partisan interests of elected officials, say the reformers, lead to oddly-shaped, and gerrymandered districts. To test this, shapes of districts in states with commissions are compared to those within the same state prior to the commission’s establishment. Additionally, shapes of districts in states with commissions are compared to those in a selected group of states without commissions. This study tests hypotheses on two methods of measuring compactness, Reock and Polsby-Popper, and coterminosity, the congruence of district lines and pre-existing political boundaries. The study finds that each state with a commission shows no significant difference in mean compactness compared to its pre-commission form. However, in aggregate, all post-commission districts show a significant increase in mean Reock compactness compared to all pre-commission districts, and all districts in states with commissions show significantly less Polsby-Popper compactness than districts in non-commission states. The study also finds no significant difference in coterminosity between commission states and non-commission states. Though the true effect of commissions may not be discernible from averages, other redistricting criteria also need to be controlled for and evaluated over time.

Understanding the causes of specific fire regimes is critical for determining the long term impacts of fire on vegetation cover. Numerous studies using 30 m Landsat data find a relationship between fire timing and vegetation type, but this relationship has not been observed at broader scales. In West Africa land-cover patterns are heterogeneous and patchy at the landscape scale and annual fires often burn mosaic patterns. It is well documented that where fires are known to be small and fragmented, the commonly used coarse-resolution MODIS data cannot give accurate estimates of burned area. Moreover, their inability to capture the spatial pattern of land-cover types burned presents a mixed pixel problem, because vegetation and agricultural fields vary on a scale less than 500 m². To overcome these issues, this study uses medium-resolution Landsat data to map land-cover. Landscape ecological indices are used to observe spatial patterns at 500 m scale.

This thesis presents a general tsunami hazard assessment for the City of Long Beach, California. Although relatively rare, tsunamis from a variety of potential sources threaten Long Beach. An anisotropic, least cost path Geographic Information Systems methodology was utilized to model approximate population exposure numbers within a number of evacuation scenarios. The variables used in the model were evacuation speed and warning time. Potential vertical evacuation sites were deduced and included within the model to compare population exposure numbers with and without the use of a vertical evacuation strategy.

The results in accordance with the literature reviewed suggest that the implementation of a vertical evacuation strategy, in addition to increased community education and preparedness, could dramatically mitigate risk and reduce the population of Long Beach's vulnerability to tsunamis, and that different areas may benefit from varying risk mitigation strategies. The implementation of vertical evacuation sites in the model decreased the population exposed by an average of 79 percent (with a mode of 99 percent).

Women entrepreneurs are on the rise and their numbers have grown at one and a half times the rate of small enterprises generally over the last 15 years. In spite of this, women are underrepresented in the fields of science, technology, engineering, and mathematics (STEM). Women face additional barriers when forging careers within these fields and obtaining startup capital. This study examines female business ownership within the fields of hi-tech and biotech, and the factors that support startups by women throughout the state of California. As both of these industry sectors are known to cluster geographically around sources of venture capital, university research and development (R&D) investment, and skilled labor, the study explored how these factors influence women entrepreneurs through two methods of analysis, specifically, a quantitative GIS analysis using exploratory geo-statistical tools, and a qualitative analysis using semi-structured interviews of twenty women business leaders. Results from the study demonstrated that factors that encourage hub formation are prone to cluster geographically, that women receive less venture capital than their male counterparts, biotech as a sector is more open than hi-tech to women’s participation, high numbers of women starting businesses alongside their alma mater, and a high participation of women in business accelerators and incubators.

The objectives of this study were to test the degree to which anthropogenic factors impact surface water geochemistry by evaluating the connections between the built environment and selected chemical parameters in the storm water runoff system for Lafayette Parish, Louisiana. An approach that combines geospatial modeling and geochemical data collected on samples from natural and concrete lined coulee drainage channels in the parish was utilized to examine (1) the distribution and interactions of different water sources (e.g. surface runoff, alternating drainage flow pathways, groundwater, etc.); and (2) the effect of impervious and agriculture land cover on water chemistry in the engineered drainage network.

Chemical compositions were analyzed for trace and major cations using an Inductively Coupled Plasma – Optical Emission Spectrometer (ICP-OES). Anion concentrations were analyzed using an Ion Chromatograph (IC). Other variables, such as dissolved oxygen, temperature, conductivity, salinity, pH, and turbidity, were measured in situ using a YSITM sonde meter. Geographic Information Systems (GIS) software was used to model drainage boundaries at each site. Impervious cover and agricultural surface layers were created from the 2010 National Land Cover database and clipped to each microwatershed drainage area such that zonal statistics could be used to calculate total area and percentages of these land coverage types in each microwatershed.

Potential groundwater interactions were evidenced to occur near the Francois Coulee Lafayette. When urban and non-urban samples were analyzed together, impervious surface cover was significantly correlated to specific conductivity (µS/cm) and the dissolved ionic species: Ca and SO4 in the microwatershed drainage areas. The exact amount of remedial acreage needed to return runoff waters to a normal baseline was calculated for targeted microwatershed drainage areas using a statistically derived coefficient estimate that suggested a 5.9 (µS/cm) increase in specific conductivity concentrations for every 1% increase in impervious surface cover.

This study performed an assessment of the representational accuracy of the forest class of the GlobeLand30 (GL30) global land cover data sets for the country of Mexico using a robust geographically distributed forest inventory survey of the forests in Mexico. The representational accuracy assessment was carried out for both the 2000 and 2010 GL30 data sets. The detailed attribute data associated with the validation set demonstrates how GL30 classifies specific forest types and how canopy coverage and number for trees per site influence the likelihood of GL30 identifying the sites correctly as forests. The results indicate that producers accuracies range from 72.3% to 97.3%. The tropical forests (89.1%) were better represented by the GL30 forest class than the temperate forest (73.9%). The most poorly represented classes from the temperate (oak: 72.3%) and tropical (low dry deciduous jungle: 74.9%) groups were deciduous. Receiver Operator Curve and Area Under the Curve analyses show that canopy coverage of a site is a better predictor of GL30, correctly identifying the site as forest for temperate forest, and that the number of the trees per site is a better predictor of GL30 correctly identifying a site as forest for tropical forests. The results also indicate a distinct spatial variability in the location of the sample sites that are misidentified as forests by GL30. The results of this thesis will help researchers and professionals better understand the representational accuracy of the GL30 data sets for the forests in Mexico.