Railroad bridges and culverts in the United States are often subject to extreme floods, which have been known to washout sections of track and ultimately lead to derailments. The potential for these events is particularly high in the western U.S. due to the lack of data, inadequate radar coverage, and the high spatial and temporal variability of storm events and terrain. In this work, a hydrologic model is developed that is capable of effectively describing the rainfall-runoff relationship of extreme thunderstorms in arid and semi-arid regions. The model was calibrated and validated using data from ten storms at the semi-arid Walnut Gulch Experimental Watershed. A methodology is also proposed for reducing the amount of raingages required to provide acceptable inputs to the hydrologic model, and also determining the most appropriate placement location for these gages. Results show that the model is capable of reproducing peak discharges, peak timings, and total volumes to within 22.1%, 12 min, and 32.8%, respectively. Results of the gage reduction procedure show that a decrease in the amount of raingages used to drive the model results in a disproportionally smaller decrease in model accuracy. Results also indicate that choosing gages using the minimization of correlation approach that is described herein will lead to an increase in model accuracy as opposed to selecting gages on a random basis.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2016-05-06T12:12:14Z No. of bitstreams: 1 emersondeoliveiramuniz.pdf: 8307586 bytes, checksum: 606af644defaa4b8c7ac04fed4b1bcbd (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2016-06-07T15:51:16Z (GMT) No. of bitstreams: 1 emersondeoliveiramuniz.pdf: 8307586 bytes, checksum: 606af644defaa4b8c7ac04fed4b1bcbd (MD5) / Made available in DSpace on 2016-06-07T15:51:16Z (GMT). No. of bitstreams: 1 emersondeoliveiramuniz.pdf: 8307586 bytes, checksum: 606af644defaa4b8c7ac04fed4b1bcbd (MD5) Previous issue date: 2013-10-31 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Numa contemporaneidade na qual as estatísticas apontam o incremento dos chamados desastres naturais em todo o mundo, sobretudo em nações como o Brasil, onde o crescente adensamento populacional nas cidades faz delas as áreas preferenciais para a realização das catástrofes, especialmente quando marcadas pela desigualdade social e pela baixa capacidade institucional de prevenção e gerenciamento desses fenômenos. A presente pesquisa faz uma interpretação geográfica do evento de inundação brusca ocorrido na cidade de Areal – RJ no dia 12 de janeiro de 2011, diretamente vinculado à catástrofe socioambiental que se processou na Região Serrana do estado naquela data. A partir da experiência vivenciada pelo próprio autor como testemunha e vítima do desastre em Areal, o trabalho registra como o evento se processou no tempo e no espaço e discute numa proposta integradora as condicionantes físicas e socioinstitucionais relacionadas ao fato da cidade ter tido a quase totalidade de sua área urbana fortemente impactada pela inundação. Dentre as condicionantes analisadas ressalta-se a presença e a operação da barragem Morro Grande, um reservatório para aproveitamento hidrelétrico a montante e próximo da área urbana do município. Duas paisagens são consideradas na interpretação, a da bacia hidrográfica do rio Piabanha e a da área urbana do município de Areal, cidade localizada no curso médio do rio que dá nome à bacia. As interações natureza-sociedade na geração do desastre, o mapeamento da área urbana afetada no evento de 2011, a análise da capacidade político-institucional de resposta à crise e o ordenamento urbano de Areal no engendramento de riscos são aspectos averiguados na pesquisa. Com base na proposta de Libaut (1971), a metodologia obedeceu quatro etapas seqüenciadas e empregou técnicas de trabalho de campo, entrevistas e mapeamento da área inundada. A pesquisa conclui que o evento na cidade foi de grande porte, deflagrado por um contexto de exceção na dinâmica flúvio-meteorológica regional, mas amplificado por fatores ligados à própria espacialidade local. / At the present times, the statistics are aiming to the increase of those called “global nature disasters”, particularly in nations like Brazil where the population of small and big cities are in constant development which contribute in addition to, the weakness of the administration and management of the prevention of disasters like the one that occurred at the city of Areal- RJ on January 12, 2011. The present work made a geographical interpretation of that event that might be the cause for that sudden flood that covered the city, located near at the highlands region of Rio de Janeiro. From that self-experience as a witness and victim the author, present a friendly proposition to discuss the construction and operation of the Morro Grande, located near that urban area and used as a Hydroelectric Reservoir to the enhancement of the region. At this point there are multiple factors to be analyzed like the area between the rivers, the urban area, the interaction nature-society, description of the area affected, the capacity political-institutional to respond to these type of events. This work is in according with Libaut (1971), this work follow the four steps and technical of work on the field. Interviews and description of the area object of this work. The paper concludes that the event in the city was large, triggered by an exception in the context of river dynamics and regional weather, but amplified by factors related to spatiality own site.

No description available.

Les crues rapides sont considérées parmi les catastrophes naturelles les plus dangereuses en raison de leur imprévisibilité et la rapidité de leurs dynamiques spatio-temporelles. Même si ces évènements touchent généralement des petites étendues spatiales, elles causent un nombre important de victimes. Les différentes analyses des impacts des crues rapides pointent la mo- bilité quotidienne comme un facteur principal augmentant l’exposition et la vulnérabilité des individus : près de la moitié des victimes sont des automobilistes surpris par les submersions des routes durant leurs trajets quotidiens. C’est dans ce contexte que plusieurs travaux ont été conduits afin de développer des outils de prévision des coupures des routes par submersion dans le département du Gard, fréquemment exposé à des évènements hydro-météorologiques extrêmes. Ces études ont permis d’identifier les points de coupure, situés aux niveaux des intersections routes/rivières, vulnérables aux crues rapides et de développer des méthodes d’estimation du risque de submersion des routes.Cette thèse s’inscrit dans la continuité de ces travaux dans le but de quantifier et de caractériser l’exposition des individus aux crues rapides durant leurs mobilités. Pour cela, nous avons contribué au développement de MobRISK : un modèle de simulation de l’ex- position des automobilistes aux submersions des routes lors d’évènements de crues rapides. Une première application de MobRISK est présentée dans le but d’évaluer l’exposition d’une population d’un zone d’étude, située au nord-ouest du Gard, à l’évènement de crue rapide des 8-9 Septembre 2002.Les résultats de ce cas d’étude ont permis d’identifier un décalage entre les dynamiques temporelles du trafic simulé au niveau des points de coupure et des niveaux de submersion de ces points. Ce décalage temporel a probablement constitué un facteur de réduction du nombre de victimes sur la route. La combinaison des estimations de submersion des routes avec le trafic simulé a permis de localiser les routes qui ont présenté le plus de danger pour les automobilistes lors de cet évènement. Ensuite, l’évaluation de l’exposition individuelle aux submersions routières a permis d’identifier les profils socio-démographiques des automobilistes les plus exposés constitués majoritairement de jeunes actifs de sexe masculin.Enfin, différents scénarios de comportements en situation de crise ont été construits et tes- tés à travers l’intégration des décisions d’adaptation des déplacements face aux perturbations environnementales. La comparaison des effets des scénarios de comportements a permis de souligner le rôle important des règles adoptées dans les prises de décision et dans l’évaluation du danger dans la réduction l’exposition des automobilistes.Au final, cette thèse souligne les potentialités et les performances de MobRISK comme un outil innovant et prometteur pour la simulation de l’exposition sociale aux crues. / Flash floods are considered as ones of the most dangerous natural hazard due to their rapidness and suddenness that leave little time for exposed people to protect themselves. Although the relatively small spatial extension of those events, several studies showed strong human impacts regarding the number of affected people. Recently, daily mobility is pointed at as important social factor increasing individual exposure and vulnerability to flash flooding : almost half of the victims are motorists trapped while travelling on flooded roads. Therefore, several studies have been conducted for assessing roads’ sensitivity to flooding in the Gard area, frequently exposed to sever flash flood events.This thesis goes a step further by integrating social dimension in order to quantify and analyze road users’ exposure to flash flood during their itineraries. It contributes to the development of MobRISK : A simulation model for assessing motorists’ exposure to roads submersion by integrating individual travel-activity patterns and behavioral adaptation regarding weather disruptions. In order to assess population exposure to September 2002 flash flood event, we conducted an application of MobRISK in a study area located in the north-ouest of the Gard department.The results show that risk of flooding is mainly located in principal road links with considerable traffic load. However, a lag time between the timing of roads’ submersion and persons crossing these roads contributes to reduce the potential vehicle-related fatal accidents. It is also found that socio-demographic variables have significant effect on individual exposure and that young working males are the most exposed persons to road flooding. Finally, different behavior scenarios are built and tested by integrating individual travel adaptation decisions regarding perceived weather disruptions. The comparison of motorists’ exposure for each scenario highlights the important role of risk attitudes and threat evaluation processes on reducing population exposure.Thus, the proposed model demonstrates the benefits of considering spatiotemporal dy- namics of population exposure to flash flood and presents an important improvement in exposure assessment methods. Such improved characterization of road users’ exposure can present valuable information for flood risk management services and emergency planning.

Els treballs d’aquesta tesi es centren geogràficament en el sector del NE de Catalunya, i concretament entre el cap de Creus i la frontera amb França. Es tracta d’una zona amb conques torrencials petites però en les que s’originen inundacions amb una certa freqüència i sovint amb danys importants. Comprèn 11 conques distribuïdes en els termes municipals de Portbou, Colera, Llançà, la Selva de Mar i el Port de la Selva. Les conques presenten unes diferències entre si, però també moltes similituds, ja que es troben en una àrea amb pluges abundants en períodes curts de precipitació. Tenen també un pendent del perfil longitudinal molt brusc, amb els cims de capçalera al voltant dels 600 m. En conseqüència s’originen avingudes sobtades o rierades (flash flood). Geològicament es tracta d’una àrea bastant homogènia, en la que la major part del territori correspon a esquistos i pissarres paleozoïques (cambroordovicianes). Hi ha un petit retall granític en el sector meridional del tram mig de la conca de Llançà. Al fons de les valls els sediments al·luvials han originat les formes recents de tipus terrassa, que aporten informació al respecte de la dinàmica hídrica dels torrents. Hi ha zones recobertes per materials antròpics de recent aportació amb diferents finalitats. Les finalitats es van basar en la cerca del coneixement de la dinàmica dels torrents, de cara a la identificació i determinació de les zones vulnerables als perills d’inundació. Dins d’aquest marc es buscava determinar la freqüència i la magnitud de les inundacions, correlacionant-les amb les característiques geomorfològiques i antròpiques del sistema fluvial. Posteriorment els objectius es van ampliar amb l’anàlisi de la incertesa ja que aparegué aquest problema, de manera general en temàtiques del medi natural i també en la mateixa zona d’estudi. Així es va plantejar determinar el perquè es produeixen aquestes incerteses (aleatòria i epistèmica) i posteriorment mirar de reduir-les. D’aquí va sorgir la necessitat d’innovar i millorar la metodologia d’identificació i cartografia de zones inundables. També es va abordar el problema del canvi climàtic. Com a base del treball es va escollir el “Mètode geomorfològic integrat d’identificació i cartografia de zones inundables” o “Mètode geomorfològic integrat”. Els primers passos de la recerca van ser l’elaboració de la cartografia geomorfològica integrada, de la qual se n’ha derivat la cartografia d’inundabilitat de cada conca. Aquestes cartografies han estat informatitzades mitjançant el programa ArcGIS, i es presenten, a escala 1:3000, en el Volum 2 d’aquest treball, i classificats per conques, començant pel N (la frontera) i acabant pel SE (cap de Creus). Els treballs de camp han estat complementats amb la recollida de dades històriques referents a esdeveniments succeïts en el passat. Amb l’aplicació del Mètode geomorfològic integrat, i a rel de l’aparició de la incertesa abans descrita, es va anar a buscar innovacions i a plantejar la problemàtica de manca de dades prou generals mitjançant l’anàlisi de tipus multiconca-multifunció. Amb el “Mètode geomorfològic integrat” es van identificar, cartografiar i analitzar els elements indicadors de zones inundables, especialment les terrasses, de les quals, en totes les conques, se'n van diferenciar dues: la T0 o llera i la T1 o superior. Les incerteses que aparegueren en l’anàlisi es centraven especialment en els càlculs dels cabals dels torrents en moments punta, però també en les determinacions dels períodes de retorn Tr de les inundacions. Les incerteses es donen, essencialment, a causa de: a) la insuficiència de dades que permetin els càlculs dels cabals i les distorsions derivades de la seva aplicació; b) la insuficiència de dades pluviomètriques; c) la insuficiència de dades en el registre històric; d) la desigualtat en la distribució de la població; e) la imperfecta predicció dels episodis d’inundació a causa de la mida de les conques; i f) la rapidesa i variabilitat de la resposta hidrològica, que fan que el temps d’actuació davant la crescuda sigui molt reduït. Pel que fa a les tipologies de zones inundables, s’ha estudiat les normatives d’àmbit europeu, estatal i autonòmic, amb la finalitat d’aplicar-les a la zona d’estudi. Aparegueren problemes d’aplicabilitat, ja que es basen en el funcionament de grans rius. Els resultats de les conques extenses i de crescuda relativament lenta no encaixaven amb els de les conques petites que tenen avingudes brusques de tipus “flash flood”. Tampoc coincidien en els períodes de retorn ni amb les zonificacions de “Zona Fluvial”, “Sistema hídric” i “Zona Inundable”. És a partir de la cartografia geomorfològica integrada i de tota la resta d’informacions que s’ha arribat a establir una classificació qualitativa de la inundabilitat: alta, mitja i baixa. És la classificació que queda recollida en la cartografia d’inundabilitat, complementada amb altres elements d’interès que hi van relacionats (zona protegida, línies de circulació preferent, punts crítics, etc.). En relació al canvi climàtic s’han buscat referències per poder preveure comportaments en futurs escenaris. S’ha buscat indrets amb possibilitats d’estudiar registres sedimentaris de materials granulomètricament fins i de naturalesa orgànica que permetessin datacions. L’anàlisi multiconca-multifunció, s’ha basat en dos aspectes principals: la magnitud de les precipitacions, i les dades històriques dels efectes de les inundacions. Ha permès la quantificació de les freqüències d’inundació, prenent els períodes dels quals es disposa de més informacions i que són més representatives. S’ha obtingut un promedi de 3 anys de temps de separació entre inundacions o període de retorn. El concepte d’inundació es considera, aquí, que les aigües desbordin de la llera del torrent ocupada per la terrassa inferior T0, i passin a ocupar l’àrea de la terrassa superior T1. En aquest concepte no es té en consideració el gruix d’aigua desbordada ni la magnitud dels efectes ni dels danys ocasionats. De tot el conjunt de treballs, ha sigut possible la formulació de conclusions referents a: - les possibilitats de zonificació de les àrees inundables, - la reducció dels factors d’incertesa característics de les conques - la introducció de millores en el Mètode geomorfològic integrat d’identificació i la cartografia de les zones inundables - la determinació de la freqüència de les inundacions mitjançant l’anàlisi multiconca-multifunció, basat en dos aspectes principals: la magnitud de les precipitacions, i les dades històriques dels efectes de les inundacions Nota: L’estructura d’aquest treball de tesi doctoral desglossa les dades en dos volums. El primer volum comprèn la totalitat dels textos, i el segon només la cartografia: la geomorfològica integrada i la d’inundabilitat, classificades per conques i amb les corresponents llegendes de simbologies gràfiques. / The study area is located in NE Catalonia, Spain, between Cap de Creus and the border with France. The area consists of eleven small basins distributed over the municipalities of Portbou, Colera, Llançà, La Selva de Mar and El Port de la Selva. These basins are prone to devastating flash floods and display differences as well as similarities given that the area is characterised by high intensity short duration rainstorms. The slope of the longitudinal profile of the basins is abrupt and the watershed attains approximately 600 m. Geologically, the area is homogeneous, consisting of palaeozoic schists and slates (cambro ordovician). A small part of the southern sector of the middle of the Llança basin is composed of granite. Recent alluvial terraces are located at the bottom of the valleys and provide valuable insights into stream dynamics. Parts of the study area are covered with diverse materials as a result of human intervention. The aim of the thesis is to improve our understanding of stream dynamics in order to identify the areas that are prone to flood hazards. An attempt was made to determine the frequency and magnitude of floods by correlating them with the geomorphological and anthropogenic characteristics of the fluvial system. Subsequent aims arose given the special characteristics of the study area, which required an analysis of uncertainty (random and epistemic). Special emphasis was placed on innovating and improving the methodology of identification and mapping of flood prone areas. Climate change was also addressed. The “Integrated geomorphological method of identification and mapping of flood hazard areas” or “Integrated geomorphological method” was used. The study consists of two parts. The first part of the study is focused on the integrated geomorphological mapping and the second part is devoted to flood hazards from which the flood hazard map for each basin was derived. These maps were computerized using ArcGis software and are at 1:3000 scale. The basins from the N (the border with France) to the E (Cap de Creus) are classified in volume 2 of the thesis. Field work was complemented by historical data of past events. The multibasin-multifunction analysis was performed to address the uncertainties that resulted from using the integrated geomorphological method. The Integrated geomorphological method enabled us to identify, map and analyse all the features providing evidence of flood prone areas especially fluvial terraces of which two were differentiated in each basin: T0 or lower terrace and T1 or upper terrace. The uncertainties of analysis affected the calculation of peak discharges and the determinations of flood frequency. These uncertainties were attributed to a) paucity of data which resulted in distortions or in the impossibility of discharge calculations, b) scarcity of rainfall data, c) insufficient historical records, d) irregular distribution of population in the basins, e) inaccurate prediction of floods because the size of the basins and f) speed and variability of the hydrological response, which considerably reduces the time for action before the flood. European, national and regional norms and zoning were studied in an effort to make a hazard classification of the flood prone areas. Problems of applicability arose because the norms are based on dynamics of large rivers. The results of the large rivers with relatively slow floods were not consistent with those of small basins that generate flash floods. These results did not coincide with those of return periods nor with the regional zoning of “Zona fluvial”, Sistema hídric” and Zona inundable” in Catalonia. The integrated geomorphological mapping together with all other data enabled us to classify qualitatively flood hazards into: high, medium and low. This classification is presented in the flood hazard mapping and is complemented by other related information (protected zone, preferential flow lines, critical points, etc.). Climate data that could predict future scenarios were obtained. It was not possible to find sedimentary records of fine grained organic sediments to enable dating. The multibasin-multifunction analysis was based on two sets of data: the magnitude of rainfall associated with flooding and historical data of floods. This enabled us to quantify the frequency of flooding by selecting the periods with more data and those that were more representative. A gap of 3 years between flooding or return period was obtained. The concept of flooding considered in the study is as follows: water overflows the lower terrace (T0) and occupies the area of the upper terrace (T1). This concept considers neither the magnitude of overflow nor the damage caused. In the light of our findings, the following conclusions may be drawn: - possibility of zoning flood prone areas - reduction of uncertainty that is characteristic of these basins - improvement in the Integrated geomorphological method of identification and mapping of flood hazard areas - determination of flood frequency by multibasin-multifunction analysis based on two main data: magnitude of rainfall and historical data (effects of floods). NB This doctoral thesis consists of 2 volumes. The first contains the text and the second includes the maps (integrated geomorphological maps and flood hazard maps) classified by basins, and the corresponding legends.

Enxurradas e inundações apresentam-se como fenômenos naturais que têm potencial de causar grandes perdas e danos à população. Estes poderiam ser reduzidos ou evitados a partir de políticas públicas de não ocupação de áreas suscetíveis ou metodologia de alerta diante da possível ocorrência destes eventos adequada. Os eventos de inundação e enxurrada são consequências de uma interação de condições físicas do terreno e de condições meteorológicas. A geomorfologia fluvial se utiliza, portanto, dos parâmetros morfométricos como ferramenta importante na descrição quantitativa das características físicas do terreno. O objetivo deste trabalho é correlacionar os dados morfométricos, extraídos a partir de dados SRTM (Shuttle Radar Topographic Mission) de sete bacias hidrográficas da Região Hidrográfica do Guaíba/Rio Grande do Sul, a episódios de inundação e enxurrada, considerando a incidência destes eventos no inventario pela Defesa Civil no período de 1991 a 2010. O modelo digital de elevação utilizado, o SRTM, possui resolução de 90m, de onde foram extraídas as variáveis analisadas neste estudo. Os resultados foram então correlacionados com os diferentes padrões de escoamento superficial das sete bacias hidrográficas. Buscando análises comparativas o estudo dividese na análise areal, linear e hipsométrica de cada um dos recortes; além de uma breve descrição das características de chuva e de uso e ocupação da área de estudo. Os resultados da análise areal, em todas as suas variáveis, mostraram-se inadequados para a tentativa de relacionar a ocorrência do evento aos seus resultados. Já os resultados obtidos pelas variáveis da análise linear mostraram-se mais eficientes na identificação de áreas de enxurrada, enquanto que a análise hipsométrica reflete melhor os episódios de inundação. Este fator de erro está provavelmente associado a dois fatores distintos: a heterogeneidade das áreas trabalhadas, por vezes generalizando regiões heterogêneas, por outras vezes supervalorizando características físicas de determinadas áreas. Além disso, credita-se erro associado aos problemas de metodologia do inventário base utilizado, este apresentando falhas no que tange a possível ausência de dados ou diferenças na obtenção e classificação dos diferentes tipos de evento. A compartimentação das bacias estudadas, visando obter recortes mais precisos e homogêneos para as diferentes áreas e uma revisão e aperfeiçoamento do inventário são apresentados como sugestões futuras para o refinamento dos resultados obtidos visando encontrar correlação entre as características morfométricas do terreno e os episódios de inundação e enxurrada. / Floods and flash floods are presented as natural phenomena that have the potential to cause major damages to the population. These could be reduced or avoided from public policies of no occupation of susceptible areas or appropriate methodology warning about the possible occurrence of the events. The events of flash flooding and flood are consequences of an interaction of an interaction of physical conditions of the terrain and weather conditions. The fluvial geomorphology uses morphometric parameters as an important tool in the quantitative description of the physical characteristics of the land. The objective of this work is to correlate the morphometric data , extracted from SRTM ( Shuttle Radar Topographic Mission ) data from seven watersheds of the Guaiba River Basin / Rio Grande do Sul , the episodes of flooding and flash flood, whereas the incidence of the incidence of these events in inventory by the Civil Defense from 1991 to 2010. The digital elevation model used, the SRTM has a resolution of 90m, where from the variables analyzed in this study were extracted. The results were then correlated with different patterns of runoff f rom seven river basins analyzed. Seeking benchmarking study is divided into areal, linear and hypsometric analysis; and a brief description of the characteristics of rain, and use and occupation of the study area were also presented. The results of areal analysis, in all its variables, proved inadequate to the attempt to relate the occurrence of the event and its results. The results obtained by the linear analysis variables were more effective in identifying areas of flash flood, while the hypsometric analysis reflects the best episodes of flooding. This error factor is probably related to two distinct factors: the heterogeneity of areas worked, sometimes generalizing heterogeneous regions, other times overestimating physical characteristics of certain areas. Also is credited to the problems of error associated inventory methodology used base, this presenting failures regarding the possible absence of data or differences in collection and classification of different types of event. The partitioning of the basins studied , aiming at obtaining more accurate and homogeneous cutouts for the various areas and a review and improvement of the inventory are presented as suggestions for future refinement of the results aimed at finding correlations between morphometric characteristics of the terrain and episodes of flooding and flash floods.

Enxurradas e inundações apresentam-se como fenômenos naturais que têm potencial de causar grandes perdas e danos à população. Estes poderiam ser reduzidos ou evitados a partir de políticas públicas de não ocupação de áreas suscetíveis ou metodologia de alerta diante da possível ocorrência destes eventos adequada. Os eventos de inundação e enxurrada são consequências de uma interação de condições físicas do terreno e de condições meteorológicas. A geomorfologia fluvial se utiliza, portanto, dos parâmetros morfométricos como ferramenta importante na descrição quantitativa das características físicas do terreno. O objetivo deste trabalho é correlacionar os dados morfométricos, extraídos a partir de dados SRTM (Shuttle Radar Topographic Mission) de sete bacias hidrográficas da Região Hidrográfica do Guaíba/Rio Grande do Sul, a episódios de inundação e enxurrada, considerando a incidência destes eventos no inventario pela Defesa Civil no período de 1991 a 2010. O modelo digital de elevação utilizado, o SRTM, possui resolução de 90m, de onde foram extraídas as variáveis analisadas neste estudo. Os resultados foram então correlacionados com os diferentes padrões de escoamento superficial das sete bacias hidrográficas. Buscando análises comparativas o estudo dividese na análise areal, linear e hipsométrica de cada um dos recortes; além de uma breve descrição das características de chuva e de uso e ocupação da área de estudo. Os resultados da análise areal, em todas as suas variáveis, mostraram-se inadequados para a tentativa de relacionar a ocorrência do evento aos seus resultados. Já os resultados obtidos pelas variáveis da análise linear mostraram-se mais eficientes na identificação de áreas de enxurrada, enquanto que a análise hipsométrica reflete melhor os episódios de inundação. Este fator de erro está provavelmente associado a dois fatores distintos: a heterogeneidade das áreas trabalhadas, por vezes generalizando regiões heterogêneas, por outras vezes supervalorizando características físicas de determinadas áreas. Além disso, credita-se erro associado aos problemas de metodologia do inventário base utilizado, este apresentando falhas no que tange a possível ausência de dados ou diferenças na obtenção e classificação dos diferentes tipos de evento. A compartimentação das bacias estudadas, visando obter recortes mais precisos e homogêneos para as diferentes áreas e uma revisão e aperfeiçoamento do inventário são apresentados como sugestões futuras para o refinamento dos resultados obtidos visando encontrar correlação entre as características morfométricas do terreno e os episódios de inundação e enxurrada. / Floods and flash floods are presented as natural phenomena that have the potential to cause major damages to the population. These could be reduced or avoided from public policies of no occupation of susceptible areas or appropriate methodology warning about the possible occurrence of the events. The events of flash flooding and flood are consequences of an interaction of an interaction of physical conditions of the terrain and weather conditions. The fluvial geomorphology uses morphometric parameters as an important tool in the quantitative description of the physical characteristics of the land. The objective of this work is to correlate the morphometric data , extracted from SRTM ( Shuttle Radar Topographic Mission ) data from seven watersheds of the Guaiba River Basin / Rio Grande do Sul , the episodes of flooding and flash flood, whereas the incidence of the incidence of these events in inventory by the Civil Defense from 1991 to 2010. The digital elevation model used, the SRTM has a resolution of 90m, where from the variables analyzed in this study were extracted. The results were then correlated with different patterns of runoff f rom seven river basins analyzed. Seeking benchmarking study is divided into areal, linear and hypsometric analysis; and a brief description of the characteristics of rain, and use and occupation of the study area were also presented. The results of areal analysis, in all its variables, proved inadequate to the attempt to relate the occurrence of the event and its results. The results obtained by the linear analysis variables were more effective in identifying areas of flash flood, while the hypsometric analysis reflects the best episodes of flooding. This error factor is probably related to two distinct factors: the heterogeneity of areas worked, sometimes generalizing heterogeneous regions, other times overestimating physical characteristics of certain areas. Also is credited to the problems of error associated inventory methodology used base, this presenting failures regarding the possible absence of data or differences in collection and classification of different types of event. The partitioning of the basins studied , aiming at obtaining more accurate and homogeneous cutouts for the various areas and a review and improvement of the inventory are presented as suggestions for future refinement of the results aimed at finding correlations between morphometric characteristics of the terrain and episodes of flooding and flash floods.

Enxurradas e inundações apresentam-se como fenômenos naturais que têm potencial de causar grandes perdas e danos à população. Estes poderiam ser reduzidos ou evitados a partir de políticas públicas de não ocupação de áreas suscetíveis ou metodologia de alerta diante da possível ocorrência destes eventos adequada. Os eventos de inundação e enxurrada são consequências de uma interação de condições físicas do terreno e de condições meteorológicas. A geomorfologia fluvial se utiliza, portanto, dos parâmetros morfométricos como ferramenta importante na descrição quantitativa das características físicas do terreno. O objetivo deste trabalho é correlacionar os dados morfométricos, extraídos a partir de dados SRTM (Shuttle Radar Topographic Mission) de sete bacias hidrográficas da Região Hidrográfica do Guaíba/Rio Grande do Sul, a episódios de inundação e enxurrada, considerando a incidência destes eventos no inventario pela Defesa Civil no período de 1991 a 2010. O modelo digital de elevação utilizado, o SRTM, possui resolução de 90m, de onde foram extraídas as variáveis analisadas neste estudo. Os resultados foram então correlacionados com os diferentes padrões de escoamento superficial das sete bacias hidrográficas. Buscando análises comparativas o estudo dividese na análise areal, linear e hipsométrica de cada um dos recortes; além de uma breve descrição das características de chuva e de uso e ocupação da área de estudo. Os resultados da análise areal, em todas as suas variáveis, mostraram-se inadequados para a tentativa de relacionar a ocorrência do evento aos seus resultados. Já os resultados obtidos pelas variáveis da análise linear mostraram-se mais eficientes na identificação de áreas de enxurrada, enquanto que a análise hipsométrica reflete melhor os episódios de inundação. Este fator de erro está provavelmente associado a dois fatores distintos: a heterogeneidade das áreas trabalhadas, por vezes generalizando regiões heterogêneas, por outras vezes supervalorizando características físicas de determinadas áreas. Além disso, credita-se erro associado aos problemas de metodologia do inventário base utilizado, este apresentando falhas no que tange a possível ausência de dados ou diferenças na obtenção e classificação dos diferentes tipos de evento. A compartimentação das bacias estudadas, visando obter recortes mais precisos e homogêneos para as diferentes áreas e uma revisão e aperfeiçoamento do inventário são apresentados como sugestões futuras para o refinamento dos resultados obtidos visando encontrar correlação entre as características morfométricas do terreno e os episódios de inundação e enxurrada. / Floods and flash floods are presented as natural phenomena that have the potential to cause major damages to the population. These could be reduced or avoided from public policies of no occupation of susceptible areas or appropriate methodology warning about the possible occurrence of the events. The events of flash flooding and flood are consequences of an interaction of an interaction of physical conditions of the terrain and weather conditions. The fluvial geomorphology uses morphometric parameters as an important tool in the quantitative description of the physical characteristics of the land. The objective of this work is to correlate the morphometric data , extracted from SRTM ( Shuttle Radar Topographic Mission ) data from seven watersheds of the Guaiba River Basin / Rio Grande do Sul , the episodes of flooding and flash flood, whereas the incidence of the incidence of these events in inventory by the Civil Defense from 1991 to 2010. The digital elevation model used, the SRTM has a resolution of 90m, where from the variables analyzed in this study were extracted. The results were then correlated with different patterns of runoff f rom seven river basins analyzed. Seeking benchmarking study is divided into areal, linear and hypsometric analysis; and a brief description of the characteristics of rain, and use and occupation of the study area were also presented. The results of areal analysis, in all its variables, proved inadequate to the attempt to relate the occurrence of the event and its results. The results obtained by the linear analysis variables were more effective in identifying areas of flash flood, while the hypsometric analysis reflects the best episodes of flooding. This error factor is probably related to two distinct factors: the heterogeneity of areas worked, sometimes generalizing heterogeneous regions, other times overestimating physical characteristics of certain areas. Also is credited to the problems of error associated inventory methodology used base, this presenting failures regarding the possible absence of data or differences in collection and classification of different types of event. The partitioning of the basins studied , aiming at obtaining more accurate and homogeneous cutouts for the various areas and a review and improvement of the inventory are presented as suggestions for future refinement of the results aimed at finding correlations between morphometric characteristics of the terrain and episodes of flooding and flash floods.

Depuis 25 ans, plus de 200 personnes ont perdu la vie dans le Sud de la France suite à une inondation (Vaison-la-Romaine en 1992, Aude en 1999, Gard en 2002, Var en 2010…). La sauvegarde des populations est affichée comme une priorité des pouvoirs publics, pourtant nous manquons encore d'une caractérisation globale de la vulnérabilité des personnes face aux phénomènes hydrométéorologiques. Ce travail de thèse postule que la mortalité liée aux catastrophes naturelles n’est pas fortuite. Elle reflète des vulnérabilités structurelles et conjoncturelles. L’analyse fine des circonstances de décès et du profil des victimes met en évidence les facteurs qui déterminent la mortalité face à de tels phénomènes et aide à mieux cibler la prévention. Ce travail s’appuie en grande partie sur la constitution d’une base de données géoréférencée pour la période 1988-2011 collectant les lieux, les circonstances et le profil des victimes. La zone d’étude couvre la France Méditerranéenne soumise aux crues « torrentielles » dans un triangle allant des Pyrénées-Orientales à l’Ardèche et au Var. L’analyse permet de remettre en perspective les préjugés d’une vulnérabilité « supposée » (femmes, enfants, personnes âgées…) sur lesquels porte de façon « réflexe » la prévention. Pour les inondations majeures (plus de dix décès) qui totalisent les 2/3 du bilan humain, les décès à domicile sont prépondérants. Les circonstances de décès laissent apparaitre une vulnérabilité subie liée à l’exposition des personnes (par exemple dans un bâti inadapté) ou une vulnérabilité structurelle liée à la capacité des personnes à résister à l’eau. A l’inverse, pour les petits évènements, moins meurtriers mais récurrents, se dégage une vulnérabilité plus « active » qui se traduit par des comportements à risque, conscients ou inconscients, notamment liés aux déplacements. La troisième partie de la thèse envisage les leviers de prévention aptes à réduire ces bilans humains. Une bonne connaissance des circonstances de décès et des facteurs expliquant la mortalité liée aux crues méditerranéennes permet d’évaluer l’efficacité et l’efficience des mesures de gestion du risque et laisse entrevoir des potentialités de « décès évitables » en ciblant les mesures de prévention adéquates. / For 25 years, more than 200 people have died in Southern France due to flooding (Vaison -la-Romaine in 1992, Aude in 1999, Gard in 2002, Var in 2010 ...). Preserving populations appears to be a government priority, however we still lack a comprehensive characterization of the vulnerability of people face to hydrometeorological phenomena. This thesis assumes that the mortality from natural disasters is not accidental. It reflects structural and circumstantial vulnerabilities. The detailed analysis of the circumstances of death and profiles of victims highlights the factors that determine mortality in this phenomena and help to improve disaster prepearedness. This work is based largely on the building of a georeferenced database for the period 1988-2011 addressing the places, the circumstances of death and the profile of the victims. The study area covers the Mediterranean France departments prone to flash floods from the Pyrenees-Orientales to the Ardeche and Var. The analysis relativizes a "supposed" vulnerability (women, childrens, elderly...). For major flooding (more than ten deaths) totaling 2/3 of losses of life, the deaths at home are most numerous. The circumstances of death let appear a sustained vulnerability to human exposure (eg. unsuitable buildings) or structural vulnerability to people's ability to resist water. In contrast, for small events, less deadly but mor recurring exudes a vulnerability more "active" which translates risk behaviors, conscious or unconscious, particularly related to displacements. The third part of the thesis considers the levers prevention able to reduce loss of life. A good knowledge of the circumstances of death and factors explaining mortality from Mediterranean floods used to assess the efficiency and effectiveness of risk management measures and suggests the potential of "avoidable deaths" targeting prevention adequate.

京都大学 / 新制・課程博士 / 博士(工学) / 甲第24212号 / 工博第5040号 / 新制||工||1787(附属図書館) / 京都大学大学院工学研究科社会基盤工学専攻 / (主査)教授 中北 英一, 准教授 山口 弘誠, 准教授 佐山 敬洋 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM