abstract: Big data that contain geo-referenced attributes have significantly reformed the way that I process and analyze geospatial data. Compared with the expected benefits received in the data-rich environment, more data have not always contributed to more accurate analysis. “Big but valueless” has becoming a critical concern to the community of GIScience and data-driven geography. As a highly-utilized function of GeoAI technique, deep learning models designed for processing geospatial data integrate powerful computing hardware and deep neural networks into various dimensions of geography to effectively discover the representation of data. However, limitations of these deep learning models have also been reported when People may have to spend much time on preparing training data for implementing a deep learning model. The objective of this dissertation research is to promote state-of-the-art deep learning models in discovering the representation, value and hidden knowledge of GIS and remote sensing data, through three research approaches. The first methodological framework aims to unify varied shadow into limited number of patterns, with the convolutional neural network (CNNs)-powered shape classification, multifarious shadow shapes with a limited number of representative shadow patterns for efficient shadow-based building height estimation. The second research focus integrates semantic analysis into a framework of various state-of-the-art CNNs to support human-level understanding of map content. The final research approach of this dissertation focuses on normalizing geospatial domain knowledge to promote the transferability of a CNN’s model to land-use/land-cover classification. This research reports a method designed to discover detailed land-use/land-cover types that might be challenging for a state-of-the-art CNN’s model that previously performed well on land-cover classification only. / Dissertation/Thesis / Doctoral Dissertation Geography 2019

Hillsborough County, Florida is subject to a variety of natural and technological hazards, which have the potential to threaten both the population and the built environment. This research focuses on several natural hazards (coastal flooding, sink hole, and hurricane) and technological hazards (toxic transportation spills and toxic release from fixed storage facilities) and the population that is potentially exposed to these hazards. Social vulnerability for this population was determined using racial composition, gender, age and household rental/ownership status. Both social vulnerability and exposure to hazardous conditions occur as a continuum across geographical space. The determination of who is exposed; the extent of exposure; and the hazardousness of their environment; requires converting this continuum into discreet values. There is little agreement on how this should be accomplished. The goal of this project is to improve on this situation by developing a multiple hazard map and a social vulnerability map using the best available data with a focus on data integration. The resulting maps were used to determine the extent that the community of Hillsborough County is exposed to hazardous conditions and the social vulnerability of that exposed community. The impact of hazard analysis is dependant on the creation of the hazard map. The hazard map can be affected by application of weighting factors to the individual or groups of hazards. Weighted linear combinations were used to examine how the exposed population changes when different hazard models are used. A technique of cumulative frequency mapping was used to examine how the composition of the exposed population changed as the hazard scores increased. This was useful in visualizing that different vulnerable communities were not exposed to hazards equally. This technique will be useful for future vulnerability/hazard assessments. The results of this research show that the most vulnerable populations in Hillsborough County, Florida are not exposed to the most extreme hazards. Instead the preponderance of the population is moderately vulnerable and is exposed to moderate hazards. It is important to focus on this population to help prepare for and respond to hazardous events and to work toward diminishing their social vulnerability.

The research presented in this Thesis is primarily concerned with the field of Geographical Information Systems (GIS) - specifically, the business applications of the technology. The empirical problem addressed is the selection of branch banks as candidates for closure using the network of branch banks of the Commonwealth Bank of Australia in the Sydney metropolitan region as the case study. Decisions to close branches are made by the Bank on the basis of performance indicators that are essentially financial. In this research, however, an alternative approach is adopted: the problem is addressed using a set of spatial criteria. Following the deregulation of the finance industry in the 1980's and the rapid introduction of new electronic channels for delivering financial services, the major banking institutions have been engaged in a process of reorganising their networks of branch banks. The most visible manifestation of this has been the ongoing and widespread closure of branches. Selecting branch banks for closure is a typical example of a complex semi-structured multi-dimensional, multi-criteria, decision-making problem. It has been well documented in previous research that Multi-Criteria Decision-Making (MCDM) models are the most appropriate ones for solving problems in this particular domain. The identification of branches for closure is also characterised by a significant spatial dimension. Decisions are based on a consideration of a number of geographical criteria and various forms of spatial analysis may be involved. An appropriate technology to assist with solving decision-making problems with a significant spatial dimension is a Spatial Decision Support System (SDSS). Most SDSS have been based on the integration of Geographical Information Systems (GIS) technology with analytical models that are proven to be best suited to specific decision-making problems and this is the approach adopted in this research. The prototype MCBC-SDSS (Multi-Criteria Branch Closure SDSS) developed here is based on the integration through the loose coupling of the ArcView GIS software with the Criterium DecisionPlus (CDP) software, which contains the suite of non-spatial analytical models that provide the analytical capability for solving multi-criteria problems. ArcView GIS is used as the engine that drives the system and to provide the analytical and display facilities to support the spatial data involved. Two MCDM models from the CDP software are used to support the decision-making analysis - the Analytical Hierarchy Process (AHP) and Simple Multi-Attribute Rating Technique (SMART). The integration of GIS with the MCDM models is based on a considerable amount of software enhancement, interface development, and computer programming. The development of the integrated system is designed to create an intelligent and user-friendly SDSS, the application of which, from the user's perspective, is a seamless operation. The success of the MCBC-SDSS is demonstrated by its application to identify candidates for closure among the 197 branches of the CBA in the Sydney metropolitan area in 2000 - the year when the building of the database for the research had been completed. The analysis is based purely on spatial considerations that have been gleaned from a major review of the literature that previous researchers have identified as affecting branch viability and performance. A set of 17 spatial variables was used as the criteria in the MCDM models. The criteria are organised in two blocks: the first includes 9 criteria relating to the characteristics of demand for branch service in the branch trade areas ('catchment area' specific criteria) while the second includes 8 criteria relating to aspects of supply provided by the existing branches in their location ('location specific' criteria). Using the developed approach, the MCBC-SDSS has been used directly to compare alternatives against criteria, not only spatial based but also financial ones, thus providing a basis for identifying the best choices regarding branch closure. The steps in the preparation of the data and the iterative procedure for implementing the MCDM models are explained and illustrated. This involves building the initial evaluation matrix, normalising the raw criteria scores, assigning weights to the criteria, and calculating priorities. Based on these, the AHP and SMART models then calculate a decision score for each branch that is used as the basis for creating the preference ranking of the branches. In this, branches with a high rank score based on the combined weighted contribution of the 17 criteria are considered to be operationally viable. On the other hand, branches with the lowest rank scores are considered as potential candidates for closure. The preference rankings generated by the models have been tested to examine their robustness in terms of the validity of criteria and their weights used in the decision analysis. Sensitivity analysis has been conducted, the results of which show that the preference rankings are stable. Different approaches have been used to validate the initial criteria, and analyse their contribution to the ranking of branch banks for closure. These help identify critical spatial variables among the 17 initial criteria selected, and suggest that some of the criteria initially selected could be deleted from the criteria list used to generate the preference rankings without substantially affecting the results. The reasonableness of the resulting preference ranking has been further demonstrated from analyses based on changing criteria weights and alternatives. The research successfully demonstrates one of the ways of enhancing the functionality of a GIS through its integration with non-spatial analytical models to develop a SDSS to aid solving decision-making problems in the selected domain. Given that to date there has been relatively few applications of SDSS similar to that developed in this research to real world decision-making problems, the procedure adopted makes it suitable for decision-making in a range of other service business applications characterised by a significant spatial dimension and multiple outlets including shopping centres, motor car dealerships, restaurant and supermarket chains. Instead of just providing solutions, however, the SDSS-based analysis in this research can better be thought of as adding value to spatial data that forms an important source of information required by decision-makers, providing insight about the situation, uncertainty, objectives, and trade-offs involved in reaching decisions, and being capable of generating alternative scenarios based on different inputs to the models that may be used to identify recommended courses of action. It can lead to better and more effective decision-making in institutions involving multi-outlet retail and service businesses and hence enables both integrated data analysis and modelling while taking multiple criteria and decision-makers' preferences into consideration.

This thesis is structured around the analysis of a model that describes the Cantonment, the Civil Lines, the Sadr Bazar and part of the Native City of the Western Indian settlement of Poona in the late nineteenth and early twentieth centuries.

<p>Since Dengue fever (DF) and its related forms, Dengue Hemorrhagic fever (DHF) and Dengue Shock Syndrome (DSS) have become important health concerns worldwide, it is also imperative to develop methods which will help in the analysis of the incidences. Dengue fever cases are growing in number as it also invades widely, affecting larger number of countries and crossing climatic boundaries. Considering that the disease as of now has neither an effective vaccine nor a cure, monitoring in order to prevent or control is the resorted alternative. GIS and its related technologies offer a wealth of interesting capabilities towards achieving this goal.</p><p><strong> </strong></p><p>The intention of this study was to develop methods to describe dengue fever outbreaks taking Rio de Janeiro, Brazil as a case study. Careful study of Census data with appropriate attributes was made to find out their potential influence on dengue fever incidence in the various regions or census districts. Dengue incidence data from year 2000 to year 2008 reported by the municipal secretariat of Rio was used to extract the necessary census districts. Base map files in MapInfo format were converted to shape files.  Using ArcGIS it was possible to merge the dengue fever incidence data with the available base map file of the City of Rio according to corresponding census districts. Choropleth maps were then created using different attributes from which patterns and trends could be used to describe the characteristic of the outbreak with respect to the socio-economic conditions. Incidence data were also plotted in Excel to see temporal variations. Cluster analysis were performed with the Moran I technique on critical periods and years of dengue outbreak. Using the square root of dengue incidence from January to April 2002 and 2008, inverse distance was selected as the conceptualised spatial relationship, Euclidean distance as the distance method. More detailed analyses were then done on the selected critical years of dengue outbreak, (years 2002 and 2008), to investigate the influence of socio-economic variables on dengue incidence per census district.</p><p> </p><p>Dengue incidence rate appeared to be higher during the rainy and warmer months between December and May. Outbreaks of dengue occurred in years 2002 and 2008 over the study period of year 2000 to 2008. Some factors included in the census data were influential in the dengue prevalence according to districts. Satisfactory results can be achieved by using this strategy as a quick method for assessing potential dengue attack, spread and possible enabling conditions. The method has the advantage where there is limited access to field work, less financial means for acquisition of data and other vital resources.</p><p>A number of difficulties were encountered during the study however and leaves areas where further work can be done for improvements. More variables would be required in order to make a complete and comprehensive description of influential conditions and factors.  There is still a gap in the analytical tools required for multi-dimensional investigations as the ones encountered in this study.  It is vital to integrate ‘GPS’ and ‘Remote Sensing’ in order to obtain a variety of up-to-date data with higher resolution.</p><p> </p>

<p>In this thesis work, we compare the results obtained from two kinds of teams forming a hierarchical organization participating in a fire fighting simulation environment called as C3Fire. First kind of teams used paper-based maps for spatial reasoning of the command tool while the other kind of teams has GIS based maps with full access to positioning data of the fire fighting units as well as sensor information about fire break.</p><p>The collected data was from 11 teams of each kind having 6 members in each team making a total of 132 participants belonging to different parts of the world.</p><p>We made a statistical analysis on the data with help of T-Test statistical medhod and a tool is designed by using Java as programming language and PostgreSQL database for importing data from log files and then applying statistical T-Test method on the fetching data from log files.The results are stored in database as well as excel files. Then a comparison is done to analyze the unit performance, communication and efficiency of both kinds of teams.</p>

<p>Command and control is central in all distributed tactical operations such as rescue operations and military operations. It takes place in a complex system of humans and artefacts, striving to reach common goals. The command and control complexity springs from several sources, including dynamism, uncertainty, risk, time pressure, feedback delays and interdependencies. Stemming from this complexity, the thesis approaches two important and related problem areas in command and control research. On a general level, the thesis seeks to approach the problems facing the command and control operators and the problems facing the designers in the associated systems development process.</p><p>We investigate the specific problem of operators losing sight of the overall perspective when working with large maps in geographical information systems with limited screen area. To approach this problem, we propose high-precision input techniques that reduce the need for zooming and panning in touch-screen systems, and informative unit representations that make better use of the screen area available. The results from an experimental study show that the proposed input techniques are as fast and accurate as state-of-the-art techniques without the need to resort to zooming. Furthermore, results from a prototype design show that the proposed unit representation reduces on-screen clutter and makes use of off-screen units to better exploit the valuable screen area.</p><p>Developing command and control systems is a complex task with several pitfalls, including getting stuck in exhaustive analyses and overrated reliance on rational methods. In this thesis, we employ a design-oriented research framework that acknowledges creative and pragmatic ingredients to handle the pitfalls. Our approach adopts the method of reconstruction and exploration of mission histories from distributed tactical operations as a means for command and control analysis. To support explorative analysis of mission histories within our framework, we propose tools for communication analysis and tools for managing metadata such as reflections, questions, hypotheses and expert comments. By using these tools together with real data from live tactical operations, we show that they can manage large amounts of data, preserve contextual data, support navigation within data, make original data easily accessible, and strengthen the link between metadata and supporting raw data. Furthermore, we show that by using these tools, multiple analysts, experts, and researchers can exchange comments on both data and metadata in a collaborative and explorative investigation of a complex scenario.</p> / Report code: LiU-Tek-Lic-2004:61.

<p>There are a number of decision making problems in which Geographical Information System (GIS) has employed to organize and facilitate the procedure of analyzing the problem. These GIS-based decision problems which typically include a number of different criteria and alternatives are generally analyzed by Multicriteria Decision Analysis (MCDA).Different locations within a geographical area represent the alternatives by which the overall goal of the project is achieved. The quality of achieving the goal is evaluated by a set of criteria which should be considered in the work. Analytic Hierarchy Process (AHP) which is a powerful method of MCDA generally can organize spatial problems and decides which alternatives are most suitable for the defined problems. However due to some intrinsic uncertainty in the method, a number of authors suggest fuzzifying the method while others are against fuzzification of the AHP.</p><p>The debate over fuzzifying AHP is going on and attempt for finding that was mostly in theory, and little, if any; practical comparison between the AHP and fuzzified AHP has done. This work presents a practical comparison of AHP and fuzzy AHP in a GIS-based problem, case study, for locating a dam in Costa Rica, considering different criteria. In order to perform the AHP and fuzzy AHP in the GIS-based problem and calculating weights of the criteria by the methods, some computer codes have written and developed in MATLAB.</p><p>The comparisons between the AHP and fuzzy AHP methods are done on result weights and on the result final maps. The comparison between the weights is repeated on different levels of uncertainty in fuzzy AHP then all the results are compared with the result of AHP method. Also this study for checking the effect of fuzzification on results is suggested Chi-Square test as a suitable tool.</p><p>Comparisons between the resulting weights of the AHP and fuzzy AHP methods show some differences between the methods. Furthermore, the Chi-Square test shows that the higher level of uncertainty in the fuzzy AHP, the greater the difference in results between the AHP and fuzzy AHP methods.</p>

This thesis presents an assessment of the variation in solifluction occurrence, morphometry and movement rates in the Abisko region, northern Sweden. Variations in movement rates are analyzed both on a regional and local scale. The main methodological contributions of this thesis have been to provide new techniques of analyzing spatial and temporal variations of solifluction in order to detect long term temporal trends and to regionalize the variations in movement rates. The spatial analysis is achieved by using a combination of field measurements, GIS and remote sensing techniques and statistical analysis. The results are presented in six papers, focusing on the morphometry of solifluction landforms (paper I), the occurrence of permafrost (paper II), the spatial and temporal variations of lobe front movement rates using aerial photographs (paper III), the temporal, regional and local spatial variations in movement rates (paper IV – VI) and statistical modelling of the occurrence of solifluction landforms and calculation of geomorphic work (paper V and VI). The results show that, on a regional scale, vegetation patterns are a major control on the occurrence of turf-banked solifluction landforms, with high NDVI-values (vegetation) associated with the presence of forms. Elevation is also a major control on a regional scale with a decrease in lobe dimensions and movement rates with increased elevation. High soil moisture values are associated with larger landforms and increased movement rates. Movement rates are generally higher in the western part of the region and appear to increase with higher MAAT. Equally, geomorphic work is greatest in the western part of the region. The important controls on a local scale vary from site to site, but include vegetation, slope angle and soil moisture. The photo analysis indicates that annual movement rates of lobe fronts in Kärkevagge and Låktatjåkka valley over the period 1959-2000 ranges from not-detectable to 63mm/yr. The permafrost model shows probabilities &gt;0.8 for permafrost at elevations above 1300 m a.s.l. in the western part of the region, decreasing to altitudes over 850 m a.s.l. in the eastern part of the region. Calculated geomorphic work suggests that solifluction is a significant denudational agent in the sub-Arctic mountains of northern Sweden, but less so than previously estimated.

This study aims to develop a method to investigate the relationship between settlement locations and the morphological landforms using geographical information systems (GIS). The method is applied to &Ccedil / ankiri province, a mountainous terrain, which is covered in seventy-seven sheets of topographic maps at 1:25.000 scale. Three databases are created and used in this study: 1) Settlement database comprising various topographic and landform attributes of 891 settlements, 2) Morphological landform database composed of 4042 landform polygon elements digitized from 1:25.000 topographic maps, and 3) Topographic database containing the digital elevation model of the area and its derivatives. The first step in the algorithm is to classify the area into four main landform classes, namely, valley, slope, flood and top. Unsuitable landforms are then clipped out based on the thresholds derived from three topographic properties (elevation, slope and aspect). Accordingly, about 2 % of the settlements and 12 % of the area are removed. The relationship is investigated using the percentages of remaining settlements and landform classes. Further analyses such as position of the settlement within the landform polygon and type of the nearest landforms are carried out for final interpretation. Following conclusions are reached on the relationship between settlement location and morphological landforms: - Percentages of settlements for flood, valley, slope and top are 8.37, 27.52, 58.60 and 5.50, respectively. Considering the percentages of the landforms provided in the area, however, valley is the most preferred landform followed by flood type. Slope and top landforms are less preferred. About 86 % of the settlements are concentrated along valley-slope boundary. - Morphological boundaries of flood and top landforms are consistent with the settlement zones. The valley-slope boundary, on the other hand, which is the most populated area, cross-cut the settlements zones.