Download (33MB) | Preview
[error in script] [error in script]
Item Type: Tesi di dottorato
Lingua: English
De Risi,
Date: 2 April 2013
Number of Pages: 200
Institution: Università degli Studi di Napoli Federico II
Department: Ingegneria strutturale
Scuola di dottorato: Ingegneria industriale
Dottorato: Ingegneria dei materiali e delle strutture
Ciclo di dottorato: 25
Coordinatore del Corso di dottorato:
Date: 2 April 2013
Number of Pages: 200
Uncontrolled Keywords: Flood, Hazard, Fragility, Exposure, Risk Assessment, Bayesian framework, Flood prone, Topographic wetness index, Urban morphology types, GIS, Informal Settlements, Africa
Settori scientifico-disciplinari del MIUR: Area 08 - Ingegneria civile e Architettura > ICAR/01 - Idraulica
Area 08 - Ingegneria civile e Architettura > ICAR/09 - Tecnica delle costruzioni
Aree tematiche (7° programma Quadro): AMBIENTE (INCLUSO CAMBIAMENTO CLIMATICO) > Cambiamenti climatici
AMBIENTE (INCLUSO CAMBIAMENTO CLIMATICO) > Proteggere i cittadini dai rischi ambientali
AMBIENTE (INCLUSO CAMBIAMENTO CLIMATICO) > Mobilitare conoscenza ambientale per la politica, l'industria e la società
Date Deposited: 08 Apr 2013 10:19
Last Modified: 15 Jul 2015 01:00


Quantification of flooding risk in urban areas is necessary as a decision-making support to stakeholders and policy-makers. It can also be effective in reducing the gap between perceived and quantified risk, in short- and long-term. The present work focuses on long-term flood risk assessment in urban areas in meso- and micro-scale. The meso-scale assessment may address an entire city; meanwhile, the micro-scale assessment covers smaller areas such as a neighborhood. This Ph.D. thesis is developed in line with the progress of the European FP7 project Climate Change and Urban Vulnerability in Africa (CLUVA). The flood risk assessment procedure is developed specifically for risk assessment in the African urban context. In the micro-scale, special attention is given to risk assessment for the informal settlements (i.e., non engineered construction). The informal settlements, which can be viewed as a direct product of rapid and un-programmed urbanization, are particularly vulnerable to flooding. They are often located in potentially flood-prone areas and are constructed without formal engineering criteria. The application of the methods adopted and developed in this work is demonstrated for various urban contexts in Africa. Meso-scale applications are presented for cities of Addis Ababa (Ethiopia), Dar es Salaam (Tanzania) and Ouagadougou (Burkina Faso). Detailed micro-scale flood risk assessment has been carried out for the urban informal settlements in the neighborhood of Suna in Dar es Salaam. A micro-scale flood hazard assessment has been performed also for the informal settlements located in the neighborhood of Little Akaki in Ethiopia. The flood risk assessment problem can be sub-divided into three main components, namely, hazard, vulnerability and exposure. Two different approaches are adopted herein for flood risk assessment in the micro- and meso-scale, respectively. The micro-scale flood hazard assessment procedure leads to the calculation of the inundation profiles --maximum flood height and velocity for each node within a lattice covering the zone of interest-- for different return periods. This is done, herein, based on both historical rainfall data and future climate projections for the precipitation. The relationship between flood height and velocity at a given point is approximated by a power-law relation. This way, only the flood height is going to be used as an interface variable for calculating flooding risk. Meso-scale flood hazard assessment in this work is based on a geo-spatial dataset of potentially flood prone urban areas, called the Topographic Wetness Index (TWI) map. In this approach, the flood-prone areas are identified by a TWI larger than a certain threshold. A GIS-based Bayesian parameter-estimation method is developed herein in order to estimate the TWI threshold based on the inundation profiles calculated for one or more micro-scale spatial windows. This Bayesian method is also used to estimate the TWI threshold based on the areal extent and delineation of flooding in previous flooding events. Flood vulnerability assessment can be regarded as the back-bone of this work. An efficient Bayesian and simulation-based algorithm is developed for the assessment of the vulnerability of a class of buildings to flooding. This Bayesian algorithm is based on assigning prescribed analytic uni- and bi-modal probability distributions for characterizing the structural fragility functions. This allows for efficient calculation of structural fragility based on a small number of (around 20-50) Monte Carlo simulations. The fragility calculations, for each of the three limit state defined in the procedure, are performed on a bi-dimensional finite-element structural model considering the openings (door and windows) constructed using the open-source software Opensees. This vulnerability assessment procedure takes into account the various sources of uncertainty due to building-to-building variability in construction geometry and detailing, and lack of complete knowledge about material mechanical properties and about characteristics of flood action (loading). The uncertain parameters considered in this work are classified into, discrete binary variables (logical statements) and continuous uncertain variables. The uncertain parameters are characterized by employing various data acquisition tools and methods, such as, orthophoto recognition, building survey, laboratory tests for material mechanical properties and literature survey. Vulnerability of the class of buildings to flooding is represented herein through the robust fragility curves, calculated as the statistics (16th, 50th and 84th percentiles) of the set of plausible fragility curves, calculated by considering the above-mentioned sources of uncertainty. Flood risk assessment in this work is developed both in meso- and micro-scale. The meso-scale flood risk assessment leads to the identification of urban flooding hot spots and the evaluation of exposure to risk. Arguably, identifying the urban flooding hot spots is one of the first steps in an integrated methodology for urban flood risk assessment and mitigation. In this work, three GIS-based datasets are employed for identifying the urban flooding hot spots for residential buildings and urban corridors (i.e., urban roads wider than 15 m). This is done by overlaying a map of potentially flood prone areas (Topographic Wetness Index, TWI), a map of residential areas and urban corridors (extracted from a city-wide assessment of urban morphology types (UMT)), and a geo-spatial Census dataset. For different statistics of the TWI threshold (e.g. MLE estimate, 16th percentile, 50th percentile), the map of the potentially flood prone areas is overlaid with the map of urban morphology units identified as residential and urban corridors in order to delineate the urban hot spots for both urban morphology types. Moreover, information related to population density is integrated by overlaying geo-spatial Census datasets in order to estimate the number of people affected by flooding. Differences in exposure characteristics are also assessed for a range of different residential types. In the micro-scale, flood risk assessment for a single-class portfolio of buildings is performed by point-wise integration of flooding hazard and the robust fragility curves. Various risk metrics are adopted in this work, namely, the mean annual frequency of exceeding a specific limit state, the annual probability of exceeding a limit state, expected number of casualties and expected replacement/reconstruction costs. The micro-scale flood risk assessment approach can be viewed as a linear and modular path starting from precipitation data (based on historical data/climate projections), to flood hazard assessment based on analysis of the hydrographic basin, to simulation-based portfolio (class) vulnerability assessment, finally leading to the assessment of flooding risk. A GIS-compatible computer platform with Matlab®-based graphical user interface is developed in this work: VISK, "Visual Vulnerability & Risk", flooding module. VISK basically mirrors the procedure for micro-scale flood risk assessment developed in this thesis. This is done by performing detailed (micro-scale) flood risk assessment for building stock with more-or-less similar characteristics. The GIS compatibility allows for graphical processing of both input and output to the program, providing an efficient visualization of flooding risk. The results can be visualized both in a detailed building-to-building scale (of potential interest to single house-holds) or as overall estimates for the entire area (of interest to policy makers).

Actions (login required)

View Item View Item