On 26 April 1986 an accident occurred in the Chernobyl nuclear power plant resulting in the release of large amount of radionuclides. Almost five percent of the total released caesium-137 was deposited in Sweden. The incidence of malignancies in the most affected counties in Sweden was investigated in three epidemiological studies. In the first study the incidence of malignancies in children and adolescents was studied for the period 1978-1992. The parishes and their inhabitants were classified according to the ground deposition of caesium-137 on an analogue map provided be the Swedish Radiological Protection Authority. A continuous increase of brain tumour incidence observed during the time of the study had no clear relationship to the Chernobyl fallout. A somewhat decreased relative risk of ALL was observed in areas with increased deposition. Other malignancies showed no changes in incidence over time or with regard to the exposure of caesium-137. In study II and III we enlarged the study base by including adults. We improved the methodology by defining a cohort of subjects who lived in the same parish from 31 December 1985 to 31 December 1987. The inhabitants from seven counties were included. Parishes were classified the same way as in study I. Due to the large number of individuals six exposure categories could be created; <3, 3–29, 30–39, 40–59, 60–79, and 80–120 kBq caesium-137/m2. The inhabitants of the 117 non-affected parishes (<3 kBq/m2) served as reference. During the 1988-1996 followup, 22,409 malignancies were recorded. The MH-IRR in the fully adjusted model was 1.00 (reference), 1.05, 1.03, 1.08, 1.10 and 1.21, respectively. ERR was 0.11 per 100 kBq/m2 (95% CL 0.03;0.20). A more advanced method was used in Study III by ignoring the exposure classification for parishes, and instead matching the dwelling coordinate to a digital map of deposition of casesium-137. In spite of a more valid exposure classification the risk estimates were similar in study II and III. Also, the ERR during the longer follow-up of 1988-1999 was almost identical, 0.10 per 100 kBq/m2 (95% CL 0.00;0.23). The strongest dose-response relationship was seen in the first four years (1988-1991). No obvious excess for leukaemia or thyroid cancer was recognised in either study II or III. The estimated number of exposure related cases was calculated to 849 in study II and 1,278 in study III. Our interpretation is that we have shown an increased incidence of total malignancies with dose-response relationship for caesium-137, only a few years after the Chernobyl accident. In study IV we compared the two different ways of classifying the exposure in study II and III. Out of the 450 parishes 111 got a different classification. The similar risk estimates in study II and III could probably be explained by relatively homogenous exposure in the parishes making the intra-parish difference less influential, especially when included in categories. In study V we examined the urinary excretion of 8-OHdG in Belarussian children from areas with high and low fallout of caesium-137, respectively. We found significantly lower urinary 8-OHdG levels in children from rural contaminated areas compared to urban uncontaminated areas, suggesting an urban, rather than a radiation related, risk factor. Using the Hill criteria for causality there is support for a causal inference between the fallout of caesium-137 from the Chernobyl accident and the increased incidence in total malignancies in Northern Sweden.

South-central Kentucky has one of the world’s most intensively studied karstareas, with most work focusing on the Mammoth Cave System and related caves and aquifers. However, slightly higher in the stratigraphic section than Mammoth Cave, the Haney Limestone is a locally important but less well studied carbonate aquifer. This research provides the most comprehensive synthesis to date of the karst hydrogeology of the Haney Limestone of south-central Kentucky, focusing on the distribution and controls on cave and karst features developed within. In contrast to drainage systems within the major limestones below, joints are the most dominant control on passage development in the Haney Limestone within the study area and the orientation of these joints is consistent with that of regional joint sets. Bedding planes and the presence of insoluble rock at the base of the Haney also exert control on conduit development in the Haney Limestone. Most of the caves of the study area developed in the Haney Limestone are singleconduit caves that receive water through direct, allogenic sources. Cave entrances are frequently perennial spring resurgences and the presence of active streams suggests that the caves function within the contemporary landscape, acting as drains for localized recharge areas. The hydrology of the Haney Limestone plays an important, if localized, role in the regional hydrology of south-central Kentucky, integrated into the current system of surface and subsurface drainage of the regional karst landscape. Evidence supports the idea that caves of the Haney Limestone are, geologically, relatively recent phenomena. A majority of the cave passages in the study area are hydrologically active, the water resurging from the sampled springs is typically undersaturated with respect to limestone, and the caves in some case appear to be developed along potential stress release fractures associated with small, apparently young valleys. This suggests that caves in the Haney Limestone were not directly influenced by the incision of the Green River over vast periods, like Mammoth Cave, but that cave development is a largely contemporary process.

This thesis presents Pixgis; a novel application for map-centric management of photographs. In short Pixgis is an interactive environment in which photographs may be discovered, viewed and managed through maps. With Pixgis finding photographs from a specific location or of a particular structure is as easy as finding the location or structure on a map. As Pixgis simultaneously displays maps, photographs and spatial metadata it also enables users to analyze photographs in new manners. This thesis work illustrates the benefits of applications for map-centric management of photographs, exposes the problems one faces when implementing such applications and presents novel solutions to many of these problems. The thesis also elaborates on spatial metadata and methods for acquisition of photographs with embedded spatial metadata.

Traffic safety is an important problem in today&rsquo / s world with increasing number of fatalities and injuries in traffic accidents. For the solution of this problem, determination of accident prone locations on a network and reasons behind is an essential step, which is studied to some extend via different traffic accident analyses in the literature. While major factors affecting accident risk, such as speed, congestion, infrastructural aspects are known, it is still very difficult to figure out the interaction among these factors, due to complexity in the spatial and temporal distribution of the aforementioned factors and traffic network characteristics. While the case of accident analysis on highways is simpler, in case of urban roads, it requires more effort due to more complex traffic networks with quite a number of conflict points and varying flow characteristics. To investigate possible relationships between speed and accident occurrence on urban arterials, a Geographic Information Systems based accident analysis methodology (GIS-TAAM) is developed in this study. This methodology uses time-dependent average link speeds (calculated from GPS-based data) and accident history of links, and three safety measures in thematic accident maps: i) total number of accident, ii) a severity index based on number of fatality and injury accidents, and iii) an alternative severity index based on total number of fatalities and injuries. The implementation of the proposed methodology and its deliverables are discussed over a pilot study on in&ouml / n&uuml / Boulevard &ndash / EskiSehir Road, Ankara.

The purpose of this study is to develop a geotechnical microzonation model regarding the suitability of the residential areas in Yenisehir (Bursa, Turkey), which is a currently developing settlement area in a seismically active region. For this purpose, soil properties and dynamic soil behaviors of the study area were assessed. Soil classification, soil amplification, natural soil predominant period, resonance phenomena and liquefaction potential of the study area were evaluated using borehole data and microtremor measurements. The raw data obtained from the previous studies carried out at Yenisehir were used for these assessments. The liquefaction potential for the study area was evaluated both in two-dimensional planimetric and three-dimensional volumetric assessments. Two geotechnical microzonation maps were produced for the study area according to the surface damage due to liquefaction (according to two different methods), soil amplification and distance to streams maps / by using Geographical Information Systems (GIS) based Multi-Criteria Decision Analysis. The weight values were assigned to the layers using Analytical Hierarchical Process method by pairwise comparisons. Evaluating geotechnical microzonation maps produced, the safest areas were found on the northern sites of the study area. The most critical areas were found to be in the middle and the southeast parts of the study area.

In der Arbeit werden Möglichkeiten des Einsatzes von Digitalen Geländemodellen (DGM) im Kontext des Hochwasserrisikomanagements aufgezeigt, wobei die Anwendung hochauflösender Laserscandaten im Fokus steht. Zunächst wird die Notwendigkeit der wissenschaftlichen Auseinandersetzung mit Hochwassergefahren hervorgehoben. Dies schließt die Betrachtung der Bedeutung von Geodaten (insbesondere DGM) im Hochwasserrisikomanagement ein. Es folgt eine Darstellung der verwendeten Daten, begleitet von einer Zusammenstellung von Methoden zur großflächigen Erfassung des Georeliefs und einer kurzen Einführung in die Hochwassermodellierung. Die wissenschaftlichen Arbeitsziele werden im Kontext eines interdisziplinären Projektes zur Hochwasserrisikoforschung definiert, in welchem DGM erstmals auch als dynamische Komponente des Risikosystems aufgefaßt werden. Die Arbeitsziele umfassen in diesem Zusammenhang die Entwicklung von automatisierten Methoden zur Gewinnung von Informationen über Deiche, insbesondere über deren genaue Lage und Höhe, und zur Modellierung von Deichgeometrien. Nach der Umsetzung und Erprobung der Verfahren in einem Geoinformationssystem wird der Prototyp einer Softwareanwendung vorgestellt, der eine automationsgestützte und nutzerfreundliche Realisierung der entwickelten Methoden erlaubt. Der Teil Deichmodellierung in Digitalen Geländemodellen beschreibt theoretische Grundlagen zur Realisierung eines speziell auf Hochwasserschutzdeiche ausgerichteten Geoinformationssystems. Es wird der Begriff des Digitalen Deichmodells (DDM) eingeführt und dessen Modellstruktur sowie Möglichkeiten zu Datenerfassung und -pflege mittels hochauflösender Laserscan-DGM erläutert. Weiterhin werden bestehende Methoden der Objekterkennung in DGM zum Zweck der Deichdetektion und -extraktion adaptiert und ein neues Verfahren entwickelt, das speziell die geometrischen Spezifika von Deichen berücksichtigt. Mit den vorgestellten Verfahren wird deren weitgehend automatisierte Erfassung und Kartierung ermöglicht. Das DDM wird derart konzipiert, daß sowohl Rückbau, Verlegung als auch die Erhöhung von Deichen im Modell umsetzbar sind. Damit wird letztlich die Simulationsfähigkeit bzw. Veränderlichkeit eines DGM als Basis für Hochwassersimulationen realisiert. Im Kapitel Vereinfachte GIS-basierte Überflutungssimulationen wird ein Konzept zur vereinfachten GIS-basierten Folgenabschätzung von Deichbrüchen vorgestellt. Dafür werden bestimmte Modellannahmen getroffen und zur Überflutungsmodellierung herangezogen. Die Ergebnisse werden im Rahmen der Modellannahmen kritisch als Möglichkeit der Abschätzung von im Extremfall zu erwartenden Wasserständen in Überschwemmungsgebieten bewertet. Nach der Dokumentation der praktischen Umsetzung der vorgestellten Verfahren werden die Ergebnisse diskutiert und Ausblicke zur nötigen bzw. möglichen weiteren Forschungsarbeiten gegeben. Der Anhang der Arbeit enthält Karten und eine Dokumentation des Programms Diketools, das als Prototyp einer Deichmodellierungssoftware implementiert wurde. / This study examines possibilities of using Digital Terrain Models (DTM) in the context of flood risk management. Thereby the focus lies on the application of high-resolution laserscan data. First the necessity of the scientific examination of flood hazards is underlined which implies the consideration of geodata (especially DTM). This is followed by an overview of methods of large-area data acquisition of the earth surface as well as a short introduction into flood modelling. The scientific objectives of this study are defined within the context of a multidisciplinary research project which for the first time considers DTM as a dynamic component in the flood risk system. The objectives resulting from this point of view comprise the development of automated methods of acquiring information on dikes. Here the exact locations of dikelines and their crest levels are of special interest. The modelling of new dike geometries and their implementiation into existing DTM data is also included. Chapter 4 (Deichmodellierung in Digitalen Geländemodellen, engl.: Dike Modelling in Digital Terrain Models) describes the theoretical fundamentals of the realisation of an information system focused on flood protection dikes. Here the term Digital Dike Model (DDM) is introduced. Accordingly its model structure is defined and possibilities of data retrieval and data management by means of high-resolution laserscan terrain models are shown. The detection and extraction of dikes is accomplished by the adaption of existing object-extraction methods. Also a new procedure has been developed which accounts for dike-specific geometrical characterisations. The presented methods enable the automated identification and mapping of dikes. The Digital Dike Model has been designed to allow the removal and relocation of dikes as well as their heightening within the model. Hereby the desired changeability of Digital Terrain Models is realised as a basis for flood simulations. In Chapter 5 (GIS-basierte Überflutungssimulationen, engl.: Simplified GIS-based Flood Modelling) a concept for a simplified estimation of dike-breach induced floodings by means of GIS-based procedures is presented. This implies the definition of certain model assumptions which are adopted when performing flood simulations. The results are discussed and the use of this method for estimating maximum flood impacts is evaluated. These parts of the thesis are followed by the discussion of the results. In the end an outlook for further research is given. The appendix of this work includes maps and a documentation of the Diketools which has been implemented as a prototype of a dike-modelling software.

Geographical Information Systems (GIS) offers preservationists a unique tool with the potential to revolutionize hazard mitigation for historic resources. The program’s ability to link information to a specific geographical location and efficiently disperse this information can solve two of the most destructive issues of current natural disaster response practices: a lack of organized information and an efficient means of disseminating this data. The resources necessary to implement a GIS program and to the requisite cooperation between both public and private preservation organizations may seem prohibitive to many preservation programs; yet, the benefits make this initial investment cost-effective. Despite efforts to mitigate disasters, both natural and man-made, their effects constantly threaten historic resources. In the past two decades, the United States has made significant strides toward a greater protection of these sites; yet damage continues to occur. In this thesis, I have investigated methods of risk mitigation implemented in the United States at both the state and local level, and in the public and private sectors, using New Orleans, Louisiana after Hurricanes Katrina and Rita as a case study. Through this analysis, I discovered that a lack of accessible, organized information and cooperation between preservationists compounded the damage caused by the actual event itself. I argue that the implementation of GIS could solve many of these issues by providing a means of both consolidating data and distributing it among responders. In this work, I demonstrate the ability of GIS to easily solve the problems of current mitigation practices for historic resources. By discussing the tools and basic functions of the program, I clearly illustrate this utility to those unfamiliar with the program, while arguing its potential as a mitigation implement to all preservationists. / text

The South African Government has set a target of universal access to basic electricity by the year 2012. Free basic electricity is defined as the amount of electricity sufficient to provide basic lighting, media access, water heating and ironing with on-grid electricity; or basic lighting and media access for a non-grid system. Eskom Distribution, in conjunction with local municipalities, is responsible for the outstanding electrification predominantly in rural areas. In KwaZulu-Natal, mountainous terrain and scattered settlement patterns of communities complicate the achievement of this goal. This study was aimed at using GIS to address the urgent need to plan electrification, firstly by identifying areas that need electrification and secondly by prioritising those areas according to set principles. Electrification areas were effectively identified and prioritised from both a need and capability of supply aspect. The study then aimed at designing the shortest networks from the grid to those identified electrification areas. To determine electrification areas spatially, electoral areas (EAs) demarcated as rural during the run up to the 1994 elections were used to identify rural areas; and Ethekwini Metropolitan Municipality, current electrification projects, reserves and a buffer zone around existing transformers excluded. Household point data was used to polygonize the remaining area, and those polygons were aggregated on their calculated area to create future rural electrification areas (FREA). A points and weighting system; based on one initially used in Namibia and further developed in an electrification planning model by RAPS Consulting, CSIR and DME to prioritise villages for electrification; was applied to calculate point scores for each FREA and other criteria such as distance from a network with capacity considered to determine a prioritised list of FREA that can be electrified immediately. Roads, land cover, household positions and slope were used to design the shortest path from the grid to the three highest scoring FREA. Each layer was reclassified, ratings applied and the layers combined to successfully determine the final path in terms of the criteria used. Interest in using GIS for spatial planning has led to a GIS Initiative Group (GISIG) being formed at Eskom Distribution in Eastern Region to address data collection, co-ordination of planning, tools written previously but never implemented being re-evaluated and, more recently, new tools being designed. However, much is still needed in terms of research, resolving of data quality issues, testing of points and weighting systems, and for functionally independent sections to work together on making changes to age-old system structures and processes before any of the recommendations resulting from this study can be effectively implemented. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2006.

The purpose of this study is to prepare a geotechnical microzonation map regarding the suitability of the residential areas in EskiSehir downtown area. In order to obtain the microzonation map, Geographical Information Systems (GIS) based Multicriteria Decision Analysis (MCDA) is used. For this analysis, the slope, flood susceptibility, soil, depth to groundwater table, swelling potential and liquefaction potential layers are prepared. The weight values to the layers and rank values to the classes of each layer are assigned by applying Simple Additive Weighting (SAW) and Analytical Hierarchical Process (AHP) methods. Two geotechnical microzonation maps are obtained as outputs of these methods. The study area is categorized into three different zones regarding the foundation suitability of residential areas as: (1) Areas suitable for settlement / (2) Provisional settlement areas / (3) Areas requiring detailed geotechnical investigations. The maps prepared using SAW and AHP methods are found to be consistent with each other. The geotechnical microzonation map prepared using AHP method is recommended as the final map of the study area.

ABSTRACT AN ATTEMPT TO CLASSIFY TURKISH DISTRICT DATA: K-MEANS AND SELF-ORGANIZING MAP (SOM) ALGORITHMS Aksoy, Ece M.S., Department of Geodetic and Geographic Information Systems Supervisor: Assoc. Prof. Dr. Oguz ISik December 2004, 112 pages There is no universally applicable clustering technique in discovering the variety of structures display in data sets. Also, a single algorithm or approach is not adequate to solve every clustering problem. There are many methods available, the criteria used differ and hence different classifications may be obtained for the same data. While larger and larger amounts of data are collected and stored in databases, there is increasing the need for efficient and effective analysis methods. Grouping or classification of measurements is the key element in these data analysis procedures. There are lots of non-spatial clustering techniques in various areas. However, spatial clustering techniques and software are not so common. This thesis is an attempt to classify Turkish district data with the help of two clustering algorithms: K-means clustering and self organizing maps (SOM). With the help of these two common techniques it is expected that a clustering can be reached, which can be used for different aims such as regional politics, constructing statistical integrity or analyzing distribution of funds, for same data in GIS environment and putting forward the facilitative usage of GIS in regional and statistical studies. All districts of Turkey, which is 923 units, were chosen as an application area in this thesis. Some limitations such as population were specified for clustering of Turkey&rsquo / s districts. Firstly, different clustering techniques for spatial classification were researched. K-Means and SOM algorithms were chosen to compare different methods with Turkey&rsquo / s district data. Afterward, database of Turkey&rsquo / s statistical datum was formed and analyzed joining with geographical data in the GIS environment. Different clustering software, ArcGIS, CrimeStat and Matlab, were applied according to conclusion of clustering techniques research. Self Organizing Maps (SOM) algorithm, which is the best and most common spatial clustering algorithm in recent years, and CrimeStat K-Means clustering were used in this thesis as clustering methods.