The prediction of motion of slowly moving landslides, also referred to as creeping slopes, is important for the reduction of landslide hazards. Such continuous and slowly moving landslides do not represent the usual stability problems of geotechnical analysis because these slopes are neither still nor ruptured but they move. For proper modeling of the motion of landslides, it is essential to develop improved techniques that integrate appropriate modeling of geological materials involved, laboratory and field tests, and verifications using computational methods. This dissertation focusses attention on the development of such an appropriate model for the time-dependent behavior of creeping landslides. Based on field observations it is proposed that the phenomenon of creeping landslides can be considered as involving the motion of a large mass of soil over a parent (fixed) mass with pronounced shear deformations occuring in a thin layer between the moving mass and the parent mass. The thin layer is refered to as interface zone while the overlying mass is refered to as solid body. The generalized Hierarchical Single Surface (HiSS) series of plasticity models are adopted to characterize the solid body. The interface zone is modeled using the specialization of the HiSS models for conditions occuring in the thin layer. Time dependency is introduced in constitutive models by adopting Perzyna's elastoviscoplastic formulation. The parameters for the HiSS and interface models are determined from laboratory tests on soils obtained from an actual slowly moving landslide at Villarbeney in Switzerland. Triaxial tests along various stress paths and oedemeter tests are conducted for the solid body. New analytical solutions are derived for prediction of oedometer tests. A general procedure for determination of viscous parameters is developed and techniques to process raw creep test data are proposed. Novel and representative simple shear interface tests are conducted to find parameters for the interface model. Special techniques for experimental analysis have been developed. A modified interface model to simulate the observed phenomenon of only compaction under shear is proposed. The parameters for the constitutive models are verified by numerically backpredicting experimental tests. An existing finite element code has been modified to incorporate various aspects of the small strain elastoviscoplastic formulation. Field measurements in the form of inclinometer profiles at various borehole locations on Villarbeney landslide are available. These inclinometer profiles are predicted using the proposed model. A comparison of the field measurements and the results from finite element analysis shows that such a model can be successfully used for predicting the behavior of slowly moving landslides.

Hazards and risks are currently identified in generic risk silos using top-down tools and methods which are incorporated into whole system risk management frameworks such as enterprise risk management. The current methods of identification and documentation are linear in approach and presentation. However, the world is multi-dimensional requiring a method of identification which responds to complex non-linear relationships. A method is required to identify cross- disciplinary hazards and formulate a register method to evidence the identified hazards. This study uses expert elicitation, web, survey and case studies to develop a method for cross-disciplinary hazard identification by application of the dimensions of generic, interface, causation and accumulation. The results of the study found many of the tools and methods used for hazard and risk identification such as hazard and operability studies took a top down approach commencing with a known failure and establishing cause and effect. The starting position of a known failure or event precludes identification of new types of failure or events and perpetuates a linear approach to hazard identification. Additionally the linear design of a risk register does not facilitate the presentation of multidimensional hazards. The current methods do not accommodate multiple lifecycles and components within cross discipline relationships. The method was applied to three case studies. The first case study had an existing risk register of 50 risks, post method application an additional 531 hazards were identified; case study (2) a register of 49 hazards and post method application additional hazards of 261; case study (3) an initial register of 45 hazards and an additional 384 hazards after method application. The impact of the method application highlights inconsistencies in the initial risk register and provides a tool which will aid the identification understanding and communication of hazards. Additionally it documents previously unidentified cross-disciplinary hazards and provides a proactive register method for identification and documentation by application of the dimensions of interface, causation and accumulation.

This thesis applies a fully nonlinear Bayesian inversion methodology to estimate shear-wave velocity (Vs) profiles and uncertainties from surface-wave dispersion data extracted from ambient seismic noise. In the inversion, the Vs profile is parameterized using a Bernstein polynomial basis, which efficiently characterizes general depth-dependent gradients in the soil/sediment column. Bernstein polynomials provide a stable parameterization in that small perturbations to the model parameters (basis-function coefficients) result in only small perturbations to the Vs profile. The inversion solution is defined in terms of the marginal posterior probability for Vs as a function of depth, estimated using Metropolis-Hastings sampling with parallel tempering. This methodology is validated via inversion of synthetic dispersion data as well as previously-considered data inverted using different parameterizations. The approach considered here is better suited than layered modelling approaches in applications where smooth gradients in geophysical parameters are expected, and/or the observed data are diffuse and not sensitive to fine-scale discrete layering (such as surface-wave dispersion). The Bernstein polynomial representation is much more general than other gradient-based models such that the form of the gradients are determined by the data, rather than by subjective parameterization choice. The Bernstein inversion methodology is also applied to dispersion data processed from passive array recordings collected in the coastal community of Kitimat, British Columbia. The region is the proposed site of several large-scale industrial development projects and has great economic and environmental significance for Canada. The inversion results are consistent with findings from other geophysical studies in the region and are used in a site-specific seismic hazard analysis. The level of ground-motion amplification expected to occur during an earthquake due to near-surface Vs structure is probabilistically quantified, and predicted to be significant compared to reference (hard ground) sites. / Graduate

Master of Arts / Department of Geography / John A. Harrington Jr / Wildcat Creek Watershed near Manhattan, Kansas, experiences damaging flash floods that have required evacuations in recent years (Spicer 2011). The purpose of this study was to qualitatively examine the issue of flooding in the Wildcat Creek Watershed through interviewing stakeholders (those that reside, own a business, or study) using a semi – structured approach. Interview discussion examined stakeholders’ perceptions of 1) how they understand the processes that create the flooding hazard, 2) whether or not they value the implementation of mitigation efforts to reduce the negative impacts of flooding, 3) whether they feel at risk to flooding, and 4) who they consider a trusted source of information about the hydrologic characteristics of the watershed. Based on the results of this study, a spatial relationship in perceptions of flooding issues in the Wildcat Creek Watershed was found. Across the study area, stakeholders understood many of the physical causes of flooding, but did not tend to see the connections among the many physical components. Overall, stakeholders believed that mitigation strategies to curb flash flooding were valuable, although many were not supportive of paying for these efforts through potential taxation from a watershed district. Despite the increase of flooding events in the past decade (Anderson 2011), many stakeholders neither saw any changes in their personal risk of exposure to flooding nor a change in their flood vulnerability. In the context of the flooding issue in Wildcat Creek Watershed, most participants trusted their neighbors and community leaders as sources of information instead of professionals who research and/or conduct work on the watershed.

L’archipel japonais jouit d’un environnement riche en ressources naturelles : ressources halieutiques, climat diversifié et relativement doux, abondance des ressources forestières, diversité de la faune et de la flore. Ces bienfaits,rendus possibles par des conditions géo-climatiques particulières, s’accompagnent d’un prix élevé. Séismes, typhons et pluies torrentielles sont le quotidien des habitants de l’archipel. Ce dernier est occasionnellement le théâtre dephénomènes plus spectaculaires comme les tsunamis et les éruptions volcaniques. Au fil des siècles, la société japonaise s’est très bien adaptée à la récurrence de ces phénomènes et aux cycles de catastrophes qu’ils induisentdans leur sillon, et elle a construit ses propres stratégies pour les surmonter, constituant un véritable paradigme cindynique. Néanmoins, au détour du XXe siècle, de nouvelles formes de danger ont émergé au sein des sociétés industrielles. Ce sont les risques technologiques, auxquels le Japon n’a pas échappé. Dans ce nouveau contexte social, politique et économique, le 11 mars 2011 un séisme et un tsunami ont frappé la centrale de Fukushima, au nord-est du Japon, et déclenché une catastrophe nucléaire d’une gravité encore difficilement sondable. Au fil des années, l’ampleur de la crise sociale, environnementale, politique et sanitaire se révélant progressivement a fait émerger de nouveaux problèmes, qui de plus en plus remettent en cause le paradigme cindynique japonais contemporain. La crise met ainsi la société japonaise à l’épreuve d’un type de catastrophe dont les enjeux collatéraux, les plus dommageables, sont également les plus difficilement appréhensibles. Cette thèse cherche à prendre de la distance par rapport à l’événement de mars 2011 tout en faisant un parallèle avec celui-ci, afin de mieux comprendre les logiques de fond qui sous-tendent le fonctionnement du paradigme cindynique « traditionnel », et ses insuffisances au regard de la situation cindynique du Japon contemporain. J’ai pour cette raison axé le centre du terrain sur la péninsule de Kii, un petit morceau de terre situé au sud de Kyōto et constitué de montagnes boisées, abritant un grand nombre de petits villages côtiers et forestiers mais aussi quelques grandes villes, exposés à des aléas très violents. À travers différents domaines du fait social et culturel, la thèse analyse les stratégies de gestion développées par les communautés de la péninsule pour vivre avec les aléas, les catastrophes et surtout les ruptures qu’ils induisent. Les angles d’approches sont multiples : l’expérience directe et la mémoire, les structures sociales, le fait religieux, les savoirs et les dimensions spatiales et temporelles du rapport aux aléas et aux événements catastrophiques. Ces différents aspects, transversaux au vécu et à la gestion des cycles catastrophiques, sont explorés conjointement pour montrer les logiques adaptatives grâce auxquelles la société japonaise a composé, dans son développement, avec les instabilités de son contexte socio-politique interne et de son milieu naturel. / Japanese archipelago has the advantage of a rich environment, full of natural resources : halieutic resources, a diversified climate, large forests blessed with a rich wildlife and flora. In the other hand, Japanese populations suffer from violent and geophysical hazards, from common earthquakes, hurricanes and huge floods to more dangerous tsunamis and volcanic eruptions. Thus, population of Japanese archipelago early developed advanced strategies to overcome the influence of natural disasters cycles over centuries, building their very own cindynic paradigm. However, in the edge of the 20th century, new danger like technologic risks rose from industrial societies like Japan’s, America’s and so on. In such changing social, economical and political context, a mega-quake, followed by a huge tsunami, stroke the northern coast of Japan, in 2011, march the 11th. The day after the tsunami hit the Fukushima nuclear power plant, a nuclear disaster alert were released by Japanese government and foreign governments around the world. The years passing by, the social, environmental, political, and health crisis that expanded from this event revealed new environmental issues, encompassing this type of disaster that redefines and question the traditional Japanese cindynic paradigm. In order to grab the dynamic underlying traditional cindynic paradigm and the reason why it revealed itself to be unable to successfully overcome the crisis that Japan faces today,I chose to move the focus to a field away from Fukushima’s nuclear issues, while keeping an eye on it. The fieldwork of this Ph.D thesis is located in Kii peninsula, a small piece of land standing south of Kyōto. Kii is the landing place of powerful tsunamis, unleashed by mega-quakes produced by the Nankai trough, every 90 to 150 year period. It is also on the usual route of violent hurricanes and torrential rains, that strike from summer to autumn, causing heavy flooding damage. This thesis approaches the problem from multiple angles, crossing exploration of direct experience, memory of disasters, knowledge, religious and political issues. Through this method, I try to enlighten underlying dynamics that made Japanese population able to live in their dangerous environment.

Inundações são eventos cada vez mais freqüentes e a cada ano os prejuízos são maiores. Inicialmente o combate dos prejuízos das inundações se deu na forma de evitá-las através de medidas estruturais. Hoje esta forma de combate vem sendo substituída, ou complementada, por medidas chamadas resilientes, geralmente medidas não estruturais, onde o objetivo é evitar as conseqüências e prejuízos da inundação, e não a própria inundação. Índices para avaliação do risco e conseqüências de inundações são ferramentas importantes para o gerenciamento do risco, eles permitem avaliar o risco, suas conseqüências, e auxiliam no desenvolvimento de planos de combate. Neste trabalho é proposto um índice de perigo (IP) e um índice de resiliência (PWRI): o primeiro lida com a vulnerabilidade de pessoas expostas à corrente de água devido inundação, e o segundo tem o objetivo de avaliar a resiliência de uma região com relação a eventos extremos hidrológicos, onde o risco de inundação e a capacidade de gerenciamento de suas conseqüências são determinados. O índice de resiliência (PWRI), que avalia qualitativamente a resiliência de uma região, é composto por seis fatores: o evento natural causador do risco (Ameaça), a probabilidade de inundação devido ao evento natural (Vulnerabilidade), a quantidade de pessoas expostas (Exposição), avaliação das medidas de prevenção e preparação contra a inundação e suas conseqüências (Antes), medidas de combate direto (Durante) e medidas de reconstrução após o evento (Depois). Os fatores são agrupados em dois grupos, o primeiro envolve a avaliação do Risco (Ameaça, Vulnerabilidade e Exposição) e o segundo a avaliação do Gerenciamento do Risco (Antes, Durante e Depois). Os fatores Ameaça, Vulnerabilidade e Exposição são obtidos a partir da previsão de escoamento superficial, modelo digital de elevação e densidade populacional. Os fatores Antes, Durante e Depois são obtidos a partir de hipóteses estabelecidas para o ano em questão. Os resultados são apresentados em uma nova forma de visualização do índice e são realizadas comparações dos resultados obtidos com eventos de inundação recentes. O índice de resiliência identifica pontos com problemas de inundações, e sugere novos pontos que podem sofrer inundação e perdas com o aumento da impermeabilização do solo e alteração do padrão de precipitação. Para o índice de perigo (IP) é desenvolvido um modelo teórico, onde os resultados são comparados e validados com dados experimentais encontrados na literatura. São apresentadas algumas recomendações e formas de utilização para o índice de perigo. Esta pesquisa forma parte do projeto temático Assessment of Impacts and Vulnerability to Climate Change in Brazil and Strategies for Adaptation Options, FAPESP 2008/58161-1 pela componente Mitigation and adaptation measures of vulnerable communities to cope with water-related risks derived from climate change scenarios at river basins of Sao Carlos. / Floods are becoming more frequent and each year the losses are greater. Initially it was used structural measures to avoid the floods and consequently avoid the losses. Today this is being replaced or supplemented by resilient measures generally in form of non-structural measures where the goal is to avoid the losses and damages and not the flood itself. Indexes for assessing the hazard and flooding losses are important tools in the risk management and allow the assessment of the flooding consequences and development of action plans. This work proposes the hazard index and the resilience index: the first one deals with the vulnerability of people exposed to flooding water flow and the second evaluates the resilience of a region with respect hydrological events. The resilience index (PWRI) is formed by six factors: a natural event which causes the threat (Hazard), the flooding probability due the hazard (Vulnerability), exposed people (Exposure), assessment of pre-flood measures (Before), emergency management (During) and post flood reconstruction measures (After). These factors are grouped into two groups: Risk assessment (Hazard, Vulnerability and Exposure) and the Risk Management (Before, During and After). The Hazard, Vulnerability and Exposure factors are obtained from the predicted runoff, digital elevation model and population density. Before, During and After factors are obtained from hypothesis. The resilience index is developed and compared with known flooding points map. It is shown that new flooding points are expected with the increase of the urbanization and land impermeabilization. The second part of the work is the hazard index (IP). A theoretical model is developed to evaluate the forces and limit conditions where people are dragged by water flow. The critical points of velocity and water depth are determined with the model and compared to experimental data found in the literature. In the end some suggestion and recommendation are present.

O objetivo deste trabalho foi identificar na cadeia produtiva do camarão Litopenaeus vannamei os perigos à saúde animal e à saúde pública e as respectivas medidas de controle, a partir de levantamentos bibliográficos. Os principais perigos à saúde animal identificados foram os vírus da Mancha Branca, Cauda Amarela, Mionecrose Infecciosa, síndrome da Taura e bactérias hepatonecrosantes dos camarões, que ocasionam grandes prejuízos para a cadeia produtiva, devido à queda da produtividade e às elevadas taxas de mortalidade. Esses perigos apresentam riscos à saúde animal, quando se mantêm viáveis nas plantas de cultivo, decorrente das práticas inadequadas de manejo sanitário e da queda de qualidade da água nos viveiros. Quanto aos riscos à saúde humana, alguns patógenos como: Vibrio spp, Salmonella spp e Staphylococcus aureus apresentam perigo se presente e viável no camarão contaminado. O uso abusivo e indiscriminado de metabissulfito de sódio em diferentes concentrações nas plantas de processamento, de contaminantes inorgânicos e resíduos de medicamentos veterinários nas criações podem resultar em níveis elevados de toxicidade aos animais e ao ambiente, precisando, consequentemente ser corrigido com a adoção de boas práticas de manejo do ambiente, implantação de vazio sanitário e estabelecimento de parâmetros microbiológicos e ambientais nos viveiros. Por ser uma atividade promissora para o desenvolvimento socioeconômico do país, é de extrema importância a adoção de medidas de controle destes perigos, que incluem a aplicação dos princípios das "Boas Práticas" e da "Análise de Perigos e Pontos Críticos de Controle" no manejo ambiental e sanitário em todos os elos da cadeia produtiva, dos criatórios até o consumo. / The objetive of this study was to identify in the production chain of shrimp Litopenaeus vannamei the hazards to animal health and public health and their control measures, from the literature. The main hazards to animal health were identified: viruses White Spot, Yellow Tail, Infectious Myonecrosis, Taura syndrome and Hepatonecrotysing bacteria of prawns, which promote large losses in the supply chain due to a drop in productivity and high mortality rates. They are considered dangerous to animal and human health, because they maintain viable pathogens in crop plants, resulting from inadequate health management practices and falling water quality in the ponds. In relation to chemical hazards, the indiscriminate use of sodium metabisulphite, inorganic contaminants and residues of veterinary drugs and pesticides can result in situations of high toxicity to animals and the environment, and must be controlled by adopting good environmental management practices, implementation of depopulation and the establishment of microbiological parameters and environmental conditions in nurseries. Physical hazards can be related to any failure in the process of industrialization and commercialization, in which the detection of foreign objects should be checked by visual inspection and the use of metal detectors. Since it is a promising socio-economic development of the country, it is extremely important to adopt measures to control these hazards, including the principles of " Good Practice" and "Hazard Analysis and Critical Control Points " in the management of environment and health in all links in the production chain, from the farms to the consumers.

Inundações são eventos cada vez mais freqüentes e a cada ano os prejuízos são maiores. Inicialmente o combate dos prejuízos das inundações se deu na forma de evitá-las através de medidas estruturais. Hoje esta forma de combate vem sendo substituída, ou complementada, por medidas chamadas resilientes, geralmente medidas não estruturais, onde o objetivo é evitar as conseqüências e prejuízos da inundação, e não a própria inundação. Índices para avaliação do risco e conseqüências de inundações são ferramentas importantes para o gerenciamento do risco, eles permitem avaliar o risco, suas conseqüências, e auxiliam no desenvolvimento de planos de combate. Neste trabalho é proposto um índice de perigo (IP) e um índice de resiliência (PWRI): o primeiro lida com a vulnerabilidade de pessoas expostas à corrente de água devido inundação, e o segundo tem o objetivo de avaliar a resiliência de uma região com relação a eventos extremos hidrológicos, onde o risco de inundação e a capacidade de gerenciamento de suas conseqüências são determinados. O índice de resiliência (PWRI), que avalia qualitativamente a resiliência de uma região, é composto por seis fatores: o evento natural causador do risco (Ameaça), a probabilidade de inundação devido ao evento natural (Vulnerabilidade), a quantidade de pessoas expostas (Exposição), avaliação das medidas de prevenção e preparação contra a inundação e suas conseqüências (Antes), medidas de combate direto (Durante) e medidas de reconstrução após o evento (Depois). Os fatores são agrupados em dois grupos, o primeiro envolve a avaliação do Risco (Ameaça, Vulnerabilidade e Exposição) e o segundo a avaliação do Gerenciamento do Risco (Antes, Durante e Depois). Os fatores Ameaça, Vulnerabilidade e Exposição são obtidos a partir da previsão de escoamento superficial, modelo digital de elevação e densidade populacional. Os fatores Antes, Durante e Depois são obtidos a partir de hipóteses estabelecidas para o ano em questão. Os resultados são apresentados em uma nova forma de visualização do índice e são realizadas comparações dos resultados obtidos com eventos de inundação recentes. O índice de resiliência identifica pontos com problemas de inundações, e sugere novos pontos que podem sofrer inundação e perdas com o aumento da impermeabilização do solo e alteração do padrão de precipitação. Para o índice de perigo (IP) é desenvolvido um modelo teórico, onde os resultados são comparados e validados com dados experimentais encontrados na literatura. São apresentadas algumas recomendações e formas de utilização para o índice de perigo. Esta pesquisa forma parte do projeto temático Assessment of Impacts and Vulnerability to Climate Change in Brazil and Strategies for Adaptation Options, FAPESP 2008/58161-1 pela componente Mitigation and adaptation measures of vulnerable communities to cope with water-related risks derived from climate change scenarios at river basins of Sao Carlos. / Floods are becoming more frequent and each year the losses are greater. Initially it was used structural measures to avoid the floods and consequently avoid the losses. Today this is being replaced or supplemented by resilient measures generally in form of non-structural measures where the goal is to avoid the losses and damages and not the flood itself. Indexes for assessing the hazard and flooding losses are important tools in the risk management and allow the assessment of the flooding consequences and development of action plans. This work proposes the hazard index and the resilience index: the first one deals with the vulnerability of people exposed to flooding water flow and the second evaluates the resilience of a region with respect hydrological events. The resilience index (PWRI) is formed by six factors: a natural event which causes the threat (Hazard), the flooding probability due the hazard (Vulnerability), exposed people (Exposure), assessment of pre-flood measures (Before), emergency management (During) and post flood reconstruction measures (After). These factors are grouped into two groups: Risk assessment (Hazard, Vulnerability and Exposure) and the Risk Management (Before, During and After). The Hazard, Vulnerability and Exposure factors are obtained from the predicted runoff, digital elevation model and population density. Before, During and After factors are obtained from hypothesis. The resilience index is developed and compared with known flooding points map. It is shown that new flooding points are expected with the increase of the urbanization and land impermeabilization. The second part of the work is the hazard index (IP). A theoretical model is developed to evaluate the forces and limit conditions where people are dragged by water flow. The critical points of velocity and water depth are determined with the model and compared to experimental data found in the literature. In the end some suggestion and recommendation are present.

As áreas urbanas precárias, também denominadas de favelas são as mais frequentemente afetadas por escorregamentos e também as mais vulneráveis, gerando danos sociais de maior monta. Este cenário resultou na demanda e no desenvolvimento de uma série de ações do poder público em diferentes municípios e estados brasileiros, culminando com o estabelecimento de um programa federal, vinculado ao Ministério das Cidades, voltado à mitigação destes riscos com o desenvolvimento de mapeamentos e a implantação de planos preventivos de defesa civil, de obras de estabilização e de reurbanização nestas áreas de risco. No estado de São Paulo, o mapeamento de escorregamento em encostas urbanas precárias vem sendo realizado desde 1990, através das iniciativas de órgãos públicos e com a participação de diversas instituições. O método de mapeamento atualmente mais utilizado emprega como técnicas principais a realização de vistorias sistemáticas de campo, investigações de superfície, utilização de fichas descritivas para armazenar as informações coletadas e a delimitação de setores de risco em imagens aéreas recentes e de grande escala e de detalhe. Este método é bastante aplicado e muitas vezes consegue subsidiar satisfatoriamente as ações de mitigação dos riscos de escorregamentos, indicando os locais prioritários para receberem estas ações. Por outro lado, o método também é questionado pelo meio técnico devido sua abordagem qualitativa, pela pouca utilização dos métodos clássicos de mapeamento geotécnico e de análise de estabilidade de taludes e pelo grau elevado de subjetividade que pode agregar, podendo produzir resultados de baixa confiabilidade. A contribuição da presente pesquisa é o aprimoramento do método de mapeamento de perigo e de risco de escorregamentos em áreas urbanas, diminuindo a subjetividade na comparação e na hierarquização dos setores, sem modificar sua abordagem fundamental e suas técnicas principais atualmente utilizadas. Para tornar o método mais sistemático, propõe-se incorporar o Processo de Análise Hierárquica (AHP) na análise dos indicadores e na hierarquização dos setores de perigo. Comparou-se o mapeamento de perigo com a aplicação do AHP com o mapeamento de risco de escorregamentos realizado em São Sebastião (SP) pelo IG-SMA em 2005, verificando-se que os resultados do mapeamento de perigo com a aplicação do AHP foram mais conservadores em algumas áreas. Entretanto, os resultados com a aplicação do AHP, diminuiu a subjetividade e evidenciou a facilidade e praticidade em se verificar a contribuição (em forma de peso) dos indicadores de perigo na classificação do perigo nos setores mapeados. Também foram analisadas as opiniões de três especialistas nos julgamentos paritários dos indicadores de perigo de escorregamentos. Os resultados não mostraram discrepâncias na classificação do perigo. / The poor urban areas, also called slums are the most frequently affected by landslides and also the most vulnerable, leading to greater social harm mounts. This situation pushed a series of actions by public authorities in several Brazilian cities and states. In the federal level was, established a program, under the auspice of the Ministry of Cities, aimed to mitigating these risks with the development of hazard and risk mapping projects and implementation of preventive plans for civil defense, works of stabilization and redevelopment in these areas of risk. In the State of São Paulo, the landslide mapping on poor and slum urban areas has been held since 1990 in several municipalities, through the initiatives of government agencies and with the participation of many institutions. The mapping method employs currently more used as main techniques to carry out systematic field surveys, surface investigations, use of fact sheets to store the information collected and the delimitation of areas of hazard and risk in aerial images, and recent large-scale and detail. This mapping method has been widely applied and often gave satisfactorily support to the initiatives to mitigate the risks of landslides, indicating the prior sites to receive those initiatives. Nevertheless, the method is questioned by the specialists because of its qualitative approach, the lack of use of traditional methods of geological and geotechnical investigation and analysis of slope stability and the high degree of subjectivity that sometimes can produce unreliable results. The expected contribution to this research is to enhance the method of hazard mapping and risk of landslides in poor and slum urban areas by reducing its subjectivity for the comparison and ranking of sectors. It is not the intent of the research to change, its fundamental approach and its main techniques currently used. To this end, it is proposed to incorporate to the method the Analytic Hierarchy Process (AHP) in the systematic analysis of the indicators and the ranking of hazard sectors. We compared the hazard mapping with the application of the AHP and landslide risk mapping conducted in São Sebastião - SP by the IG-SMA in 2005, verifying that the results of the hazard mapping in the application of the AHP were more conservative in some areas. However, the results from the application of the AHP, decreased the subjectivity and demonstrated the ease and convenience to verify the contribution (in the form of weight) of the hazard indicators in the ranking of hazard in the areas mapped. We also analyzed the opinions of three experts in the trials of parity landslides hazard indicators. The results showed no discrepancies in the ranking of hazard.

O aumento da população em grandes centros urbanos, dos problemas ambientais, e as mudanças climáticas globais são alguns dos fatores que têm influenciado numa maior intensidade e a frequência de eventos extremos e de desastres naturais em todo o globo. A região sul do Brasil aparece como uma das mais afetadas, sofrendo com estiagens e inundações recorrentes. O presente estudo tem como objetivo analisar o risco de inundação da bacia hidrográfica do arroio Feijó que sofre frequentemente com inundações. Baseia-se na abordagem metodológica do Índice de Vulnerabilidade, que contempla dados demográficos e sociais dos setores censitários (IBGE). Abrange áreas dos municípios de Porto Alegre, Alvorada e Viamão, no estado do Rio Grande do Sul. Ao longo do período analisado, de 1980 a 2015, foram registradas 36 inundações em 19 anos, com a probabilidade de 53% ocorrência de um novo evento danoso a cada ano. Os meses com maior frequência são fevereiro setembro e outubro. Ainda que 63% dos anos com inundações tenham apontado atuação do fenômeno El Niño – Oscilação Sul, não é possível apontar influência direta dos mesmos nesses processos. Entendendo o risco como a relação do perigo ou ameaça de ocorrência de eventos adversos com a vulnerabilidade dos elementos expostos gerou-se a espacialização desses fatores e o mapa de risco de inundação. O perigo expressa a probabilidade quantitativa ou qualitativa de recorrência, abrangência, intensidade que ocorre o processo. Por meio da mancha de inundação de tempo de retorno de 100 anos, obteve-se o mapa de perigo. Os setores apontados com alto a muito alto perigo localizam-se nos bairros de Alvorada (Passo do Feijó, Umbu, Jardim Algarve e Americana) e Porto Alegre (Rubem Berta), apresentando de 50% a 80% das suas áreas com possibilidade de serem inundadas. O levantamento por meio do índice de vulnerabilidade proposto apontou os setores mais vulneráveis de toda a bacia em geral, bem como os que indicam qual população está mais exposta aos eventos adversos. Os municípios de Alvorada e Viamão foram os que apontaram mais setores com índice de elevado de vulnerabilidade. Setores que apontam baixo desenvolvimento social e baixa capacidade de enfretamento aos desastres. Por meio do mapa de risco, foram identificadas as áreas com baixo a muito alto riscos de inundação da BHAF, os setores censitários mais críticos, encontram-se nos bairros Passo do Feijó e Americana em Alvorada. A análise de risco demostra ser importante ferramenta para identificar áreas que necessitem maior atenção do poder público e investimentos em ações de proteção, prevenção e mitigação. A metodologia demonstrou ser eficaz, sendo de fácil execução, implementação e baixo custo. Pretende-se dessa forma, contribuir através desse estudo para uma gestão de riscos e desastres mais eficiente na área. / The increase in population in large urban centers, enviromental issues and global climate changes, among others factors, have increased the frequency and the intensity of extreme events and natural disasters across the globe and in this country. The south appears to be the most affected region, suffering from droughts and floods. This study aims to analyze the risks of flooding at the Feijo stream basin which often suffers from floods. The study is based on the methodological approach of the vulnerability index, which includes demographic and social data from the census tracts (IBGE). It covers areas of the municipalities of Porto Alegre, Alvorada e Viamão, at the Rio Grande do Sul state. Over the period analyzed, from 1980 to 2015, 36 floods were recorded in 19 years, with the probability of a 53% occurrence of a new damaging event each year. The most frequent months are February, September and October. Although 63% of the years with the occurrence of floods have also shown the occurrence of El Niño - South Oscillation phenomenon, it is not possible to indicate direct influence. Recognizing risk as the relation of danger or threat of adverse occurring events with the vulnerability of the elements exposed is how the spatialization of these factors and the flood risk map were generated. The hazard expresses the quantitative or qualitative probability of recurrence, extent, intensity in which the process occurs. By means of the 100-year return time flood spot the hazard map was obtained. The high-risk areas are located in the Alvorada (Passo do Feijó, Umbu, Jardim Algarve and Americana) and Porto Alegre (Rubem Berta) neighborhoods, pointing to 50% to 80% of these areas with the possibility of being flooded. The survey using the proposed Vulnerability index showed the most vulnerable sectors of the entire basin, as well as which population groups are most exposed to adverse events. The municipalities of Alvorada and Viamão were those that indicated more sectors with a high index of vulnerability. These are sectors that show low social development and low ability to cope with disasters. By means of the risk map, the areas with low to very high flood risks of BHAF (the most critical census tracts) were identified in the Passo do Feijó and Americana neighborhoods in Alvorada. The risk analysis proves to be an important tool to identify areas that need greater public attention and investments in protection, prevention and mitigation actions. The methodology proved to be effective, being to put into effect and implement, and at a low cost. The aim is to contribute through this study to a more efficient risk and disaster management in the area.