scholarly journals Testing empirical and synthetic flood damage models: the case of Italy

2019 ◽  
Vol 19 (3) ◽  
pp. 661-678 ◽  
Author(s):  
Mattia Amadio ◽  
Anna Rita Scorzini ◽  
Francesca Carisi ◽  
Arthur H. Essenfelder ◽  
Alessio Domeneghetti ◽  
...  
Keyword(s):  
Risk Management ◽  
Flood Hazard ◽  
Residential Buildings ◽  
Flood Damage ◽  
Machine Learning Techniques ◽  
Flood Events ◽  
Ancillary Data ◽  
Data Set ◽  
Damage Models ◽  
Multivariable Models

Abstract. Flood risk management generally relies on economic assessments performed by using flood loss models of different complexity, ranging from simple univariable models to more complex multivariable models. The latter account for a large number of hazard, exposure and vulnerability factors, being potentially more robust when extensive input information is available. We collected a comprehensive data set related to three recent major flood events in northern Italy (Adda 2002, Bacchiglione 2010 and Secchia 2014), including flood hazard features (depth, velocity and duration), building characteristics (size, type, quality, economic value) and reported losses. The objective of this study is to compare the performances of expert-based and empirical (both uni- and multivariable) damage models for estimating the potential economic costs of flood events to residential buildings. The performances of four literature flood damage models of different natures and complexities are compared with those of univariable, bivariable and multivariable models trained and tested by using empirical records from Italy. The uni- and bivariable models are developed by using linear, logarithmic and square root regression, whereas multivariable models are based on two machine-learning techniques: random forest and artificial neural networks. Results provide important insights about the choice of the damage modelling approach for operational disaster risk management. Our findings suggest that multivariable models have better potential for producing reliable damage estimates when extensive ancillary data for flood event characterisation are available, while univariable models can be adequate if data are scarce. The analysis also highlights that expert-based synthetic models are likely better suited for transferability to other areas compared to empirically based flood damage models.

scholarly journals Estimating flood damage in Italy: empirical vs expert-based modelling approach

2018 ◽  
Author(s):  
Mattia Amadio ◽  
Anna Rita Scorzini ◽  
Francesca Carisi ◽  
Arthur H. Essenfelder ◽  
Alessio Domeneghetti ◽  
...  
Keyword(s):  
Risk Management ◽  
Flood Hazard ◽  
Residential Buildings ◽  
Economic Value ◽  
Flood Damage ◽  
Machine Learning Techniques ◽  
Flood Events ◽  
Damage Models ◽  
Multivariable Models ◽  
Modelling Approach

Abstract. Flood risk management generally relies on economic assessments performed using flood loss models of different complexity, ranging from simple univariable to more complex multivariable models. These latter accounts for a large number of hazard, exposure and vulnerability factors, being potentially more robust when extensive input information is available. In this paper we collected a comprehensive dataset related to three recent major flood events in Northern Italy (Adda 2002, Bacchiglione 2010 and Secchia 2014), including flood hazard features (depth, velocity and duration), buildings characteristics (size, type, quality, economic value) as well as reported losses. The objective of this study is to compare the performances of expert-based and empirical (both uni- and multivariable) damage models for estimating the potential economic costs of flood events to residential buildings. The performance of four literature flood damage models of different nature and complexity are compared with the performance of univariable, bivariable and multivariable models empirically developed for Italy and tested at the micro scale based upon observed records. The uni- and bivariable models are produced testing linear, logarithmic and square root regression while multivariable models are based on two machine learning techniques, namely Random Forest and Artificial Neural Networks. Results provide important insights about the choice of the damage modelling approach for operational disaster risk management.

scholarly journals Development and assessment of uni- and multi-variable flood loss models for Emilia-Romagna (Italy)

2017 ◽  
Author(s):  
Francesca Carisi ◽  
Kai Schröter ◽  
Alessio Domeneghetti ◽  
Heidi Kreibich ◽  
Attilio Castellarin
Keyword(s):  
Residential Buildings ◽  
Economic Value ◽  
Flood Damage ◽  
Data Sets ◽  
Local Data ◽  
Data Set ◽  
Damage Models ◽  
Flood Loss ◽  
Loss Data ◽  
Loss Models

Abstract. Simplified flood loss models are one important source of uncertainty in flood risk assessments. Many countries experience sparseness or absence of comprehensive high-quality flood loss data sets which is often rooted in a lack of protocols and reference procedures for compiling loss data sets after flood events. Such data are an important reference for developing and validating flood loss models. We consider the Secchia river flood event of January 2014, when a sudden levee-breach caused the inundation of nearly 52 km2 in Northern Italy. For this event we compiled a comprehensive flood loss data set of affected private households including buildings footprint, economic value, damages to contents, etc. based on information collected by local authorities after the event. By analysing this data set we tackle the problem of flood damage estimation in Emilia-Romagna (Italy) by identifying empirical uni- and multi-variable loss models for residential buildings and contents. The accuracy of the proposed models is compared with those of several flood-damage models reported in the literature, providing additional insights on the transferability of the models between different contexts. Our results show that (1) even simple uni-variable damage models based on local data are significantly more accurate than literature models derived for different contexts; (2) multi-variable models that consider several explanatory variables outperform uni-variable models which use only water depth. However, multi-variable models can only be effectively developed and applied if sufficient and detailed information is available.

scholarly journals Development and assessment of uni- and multivariable flood loss models for Emilia-Romagna (Italy)

2018 ◽  
Vol 18 (7) ◽  
pp. 2057-2079 ◽  
Author(s):  
Francesca Carisi ◽  
Kai Schröter ◽  
Alessio Domeneghetti ◽  
Heidi Kreibich ◽  
Attilio Castellarin
Keyword(s):  
Residential Buildings ◽  
Flood Damage ◽  
Local Data ◽  
Explanatory Variables ◽  
Damage Models ◽  
Private Households ◽  
Flood Loss ◽  
Multivariable Models ◽  
Loss Models ◽  
Flood Damage Estimation

Abstract. Flood loss models are one important source of uncertainty in flood risk assessments. Many countries experience sparseness or absence of comprehensive high-quality flood loss data, which is often rooted in a lack of protocols and reference procedures for compiling loss datasets after flood events. Such data are an important reference for developing and validating flood loss models. We consider the Secchia River flood event of January 2014, when a sudden levee breach caused the inundation of nearly 52 km2 in northern Italy. After this event local authorities collected a comprehensive flood loss dataset of affected private households including building footprints and structures and damages to buildings and contents. The dataset was enriched with further information compiled by us, including economic building values, maximum water depths, velocities and flood durations for each building. By analyzing this dataset we tackle the problem of flood damage estimation in Emilia-Romagna (Italy) by identifying empirical uni- and multivariable loss models for residential buildings and contents. The accuracy of the proposed models is compared with that of several flood damage models reported in the literature, providing additional insights into the transferability of the models among different contexts. Our results show that (1) even simple univariable damage models based on local data are significantly more accurate than literature models derived for different contexts; (2) multivariable models that consider several explanatory variables outperform univariable models, which use only water depth. However, multivariable models can only be effectively developed and applied if sufficient and detailed information is available.

scholarly journals The object-specific flood damage database HOWAS 21

2020 ◽  
Vol 20 (9) ◽  
pp. 2503-2519
Author(s):  
Patric Kellermann ◽  
Kai Schröter ◽  
Annegret H. Thieken ◽  
Sören-Nils Haubrock ◽  
Heidi Kreibich
Keyword(s):  
Web Application ◽  
Flood Hazard ◽  
Flood Damage ◽  
Economic Sectors ◽  
Hazard Exposure ◽  
Damage Models ◽  
Large Potential ◽  
Damage Data ◽  
Flood Analysis

Abstract. The Flood Damage Database HOWAS 21 contains object-specific flood damage data resulting from fluvial, pluvial and groundwater flooding. The datasets incorporate various variables of flood hazard, exposure, vulnerability and direct tangible damage at properties from several economic sectors. The main purpose of development of HOWAS 21 was to support forensic flood analysis and the derivation of flood damage models. HOWAS 21 was first developed for Germany and currently almost exclusively contains datasets from Germany. However, its scope has recently been enlarged with the aim to serve as an international flood damage database; e.g. its web application is now available in German and English. This paper presents the recent advancements of HOWAS 21 and highlights exemplary analyses to demonstrate the use of HOWAS 21 flood damage data. The data applications indicate a large potential of the database for fostering a better understanding and estimation of the consequences of flooding.

scholarly journals Are OpenStreetMap building data useful for flood vulnerability modelling?

2021 ◽  
Vol 21 (2) ◽  
pp. 643-662
Author(s):  
Marco Cerri ◽  
Max Steinhausen ◽  
Heidi Kreibich ◽  
Kai Schröter
Keyword(s):  
Data Model ◽  
Mean Squared Error ◽  
Residential Buildings ◽  
Residential Building ◽  
Predictive Performance ◽  
Flood Damage ◽  
Model Learning ◽  
Data Set ◽  
Flood Vulnerability ◽  
Spatial Measures

Abstract. Flood risk modelling aims to quantify the probability of flooding and the resulting consequences for exposed elements. The assessment of flood damage is a core task that requires the description of complex flood damage processes including the influences of flooding intensity and vulnerability characteristics. Multi-variable modelling approaches are better suited for this purpose than simple stage–damage functions. However, multi-variable flood vulnerability models require detailed input data and often have problems in predicting damage for regions other than those for which they have been developed. A transfer of vulnerability models usually results in a drop of model predictive performance. Here we investigate the questions as to whether data from the open-data source OpenStreetMap is suitable to model flood vulnerability of residential buildings and whether the underlying standardized data model is helpful for transferring models across regions. We develop a new data set by calculating numerical spatial measures for residential-building footprints and combining these variables with an empirical data set of observed flood damage. From this data set random forest regression models are learned using regional subsets and are tested for predicting flood damage in other regions. This regional split-sample validation approach reveals that the predictive performance of models based on OpenStreetMap building geometry data is comparable to alternative multi-variable models, which use comprehensive and detailed information about preparedness, socio-economic status and other aspects of residential-building vulnerability. The transfer of these models for application in other regions should include a test of model performance using independent local flood data. Including numerical spatial measures based on OpenStreetMap building footprints reduces model prediction errors (MAE – mean absolute error – by 20 % and MSE – mean squared error – by 25 %) and increases the reliability of model predictions by a factor of 1.4 in terms of the hit rate when compared to a model that uses only water depth as a predictor. This applies also when the models are transferred to other regions which have not been used for model learning. Further, our results show that using numerical spatial measures derived from OpenStreetMap building footprints does not resolve all problems of model transfer. Still, we conclude that these variables are useful proxies for flood vulnerability modelling because these data are consistent (i.e. input variables and underlying data model have the same definition, format, units, etc.) and openly accessible and thus make it easier and more cost-effective to transfer vulnerability models to other regions.

What is the value of a multi-decadal regional database of compensation claims for flood damage modelling and risk management?

2021 ◽  
Author(s):  
Axelle Doppagne ◽  
Pierre Archambeau ◽  
Jacques Teller ◽  
Anna Rita Scorzini ◽  
Daniela Molinari ◽  
...  
Keyword(s):  
Flood Damage ◽  
State Agency ◽  
Flow Depth ◽  
Flood Events ◽  
Explanatory Variables ◽  
Damage Models ◽  
Damage Modelling ◽  
Spatial Coverage ◽  
Compensation Claims ◽  
Building Level

<p>Flood damage modelling is a key component of flood risk modelling, assessment and management. Reliable empirical data of flood damage are essential to support the development and validation of flood damage models. However, such datasets remain scarce and incomplete, particularly those combining a large spatial coverage (e.g., regional, national) over a long time period (e.g., several decades) with a detailed resolution (e.g., address-level data).</p><p>In this research, we analysed a database of 27,000 compensation claims submitted to a Belgian state agency (Disaster Fund). It covers 104 natural disasters of various types (incl. floods, storms, rockslides …) which occurred in the Walloon region in Belgium between 1993 and 2019. The region extends over parts of the Meuse and of the Scheldt river basins. The registered amounts of damage at the building level were estimated by state-designated experts. They are classified in six categories. While roughly half of the registered disasters are pluvial flooding events, they account for less than a quarter of the total claimed damage. In contrast, riverine floods correspond to about one third of the registered events, but they lead to one half of the claimed damage.</p><p>A detailed analysis of the data was undertaken for a limited number of major riverine flood events (1993, 1995, 2002), which have caused a very large portion of the total damage. By geo-referencing the postal address of each individual building, it was possible to assign each claim to a specific river reach. This enabled pointing at the most flood prone river stretches in an objective way. Then, using cadastral data, each type and amount of damage could be attributed to a specific building.</p><p>To explore the value of the database for elaborating and validating damage models, the claimed damage data at the building level were related to estimates of hydraulic variables for the corresponding flood events. To do so, we used an existing database of results of 2D hydrodynamic modelling, covering 1,200+ km of river reaches and providing raster files at a spatial resolution ranging from 2 m to 5 m for computed flow depth and velocity in the floodplains. The attribution of flow depth to individual buildings was performed either by averaging the computed flow depths around the building footprint or by considering the maximum value.</p><p>The correlation between claimed damage at the building level and attributed flow depth is relatively low, irrespective of the flow depth attribution method. This may result from the high uncertainty affecting each of these variables. It also hints at the necessity of using multivariable damage models which account for a broader range of explanatory variables than the sole flow depth (flow velocity, characteristics of building material and equipment, building age, etc.). This will be discussed in the presentation and further explored in the next steps of this research.</p><p>Data for this analysis were provided by the Belgian regional agency SPW-IAS in July 2020. Due to privacy reasons, data at the address-level may not be disseminated in the scientific community; but results of data processing may be shared at an aggregated level.</p>

scholarly journals Flood fatality hazard and flood damage hazard: combining multiple hazard characteristics into meaningful maps for spatial planning

2015 ◽  
Vol 15 (6) ◽  
pp. 1297-1309 ◽  
Author(s):  
K. M. de Bruijn ◽  
F. Klijn ◽  
B. van de Pas ◽  
C. T. J. Slager
Keyword(s):  
Risk Management ◽  
Flood Risk ◽  
Spatial Planning ◽  
Flood Hazard ◽  
Flood Damage ◽  
Flood Risk Management ◽  
Mitigation Measures ◽  
Accurate Information ◽  
Hazard Maps ◽  
The Past

Abstract. For comprehensive flood risk management, accurate information on flood hazards is crucial. While in the past an estimate of potential flood consequences in large areas was often sufficient to make decisions on flood protection, there is currently an increasing demand to have detailed hazard maps available to be able to consider other risk-reducing measures as well. Hazard maps are a prerequisite for spatial planning, but can also support emergency management, the design of flood mitigation measures, and the setting of insurance policies. The increase in flood risks due to population growth and economic development in hazardous areas in the past shows that sensible spatial planning is crucial to prevent risks increasing further. Assigning the least hazardous locations for development or adapting developments to the actual hazard requires comprehensive flood hazard maps. Since flood hazard is a multi-dimensional phenomenon, many different maps could be relevant. Having large numbers of maps to take into account does not, however, make planning easier. To support flood risk management planning we therefore introduce a new approach in which all relevant flood hazard parameters can be combined into two comprehensive maps of flood damage hazard and flood fatality hazard.

scholarly journals Surface water floods in Switzerland: what insurance claim records tell us about the damage in space and time

2017 ◽  
Vol 17 (9) ◽  
pp. 1659-1682 ◽  
Author(s):  
Daniel B. Bernet ◽  
Volker Prasuhn ◽  
Rolf Weingartner
Keyword(s):  
Surface Water ◽  
Flood Hazard ◽  
Geographical Location ◽  
Flood Damage ◽  
Flood Hazards ◽  
Data Set ◽  
Space And Time ◽  
Alpine Regions ◽  
Spatial And Temporal Distributions ◽  
Almost All

Abstract. Surface water floods (SWFs) have received increasing attention in the recent years. Nevertheless, we still know relatively little about where, when and why such floods occur and cause damage, largely due to a lack of data but to some degree also because of terminological ambiguities. Therefore, in a preparatory step, we summarize related terms and identify the need for unequivocal terminology across disciplines and international boundaries in order to bring the science together. Thereafter, we introduce a large (n = 63 117), long (10–33 years) and representative (48 % of all Swiss buildings covered) data set of spatially explicit Swiss insurance flood claims. Based on registered flood damage to buildings, the main aims of this study are twofold: First, we introduce a method to differentiate damage caused by SWFs and fluvial floods based on the geographical location of each damaged object in relation to flood hazard maps and the hydrological network. Second, we analyze the data with respect to their spatial and temporal distributions aimed at quantitatively answering the fundamental questions of how relevant SWF damage really is, as well as where and when it occurs in space and time. This study reveals that SWFs are responsible for at least 45 % of the flood damage to buildings and 23 % of the associated direct tangible losses, whereas lower losses per claim are responsible for the lower loss share. The Swiss lowlands are affected more heavily by SWFs than the alpine regions. At the same time, the results show that the damage claims and associated losses are not evenly distributed within each region either. Damage caused by SWFs occurs by far most frequently in summer in almost all regions. The normalized SWF damage of all regions shows no significant upward trend between 1993 and 2013. We conclude that SWFs are in fact a highly relevant process in Switzerland that should receive similar attention like fluvial flood hazards. Moreover, as SWF damage almost always coincides with fluvial flood damage, we suggest considering SWFs, like fluvial floods, as integrated processes of our catchments.

scholarly journals Review article: A review and critical analysis of the efforts towards urban flood risk management in the Lagos region of Nigeria

2016 ◽  
Vol 16 (2) ◽  
pp. 349-369 ◽  
Author(s):  
U. C. Nkwunonwo ◽  
M. Whitworth ◽  
B. Baily
Keyword(s):  
Risk Management ◽  
Flood Risk ◽  
Risk Mitigation ◽  
Flood Hazard ◽  
Flood Management ◽  
Flood Damage ◽  
Flood Risk Management ◽  
Management Policy ◽  
Urban Flooding ◽  
Urban Flood

Abstract. Urban flooding has been and will continue to be a significant problem for many cities across the developed and developing world. Crucial to the amelioration of the effects of these floods is the need to formulate a sound flood management policy, which is driven by knowledge of the frequency and magnitude of impacts of these floods. Within the area of flood research, attempts are being made to gain a better understanding of the causes, impacts, and pattern of urban flooding. According to the United Nations office for disaster reduction (UNISDR), flood risk is conceptualized on the basis of three integral components which are frequently adopted during flood damage estimation. These components are: probability of flood hazard, the level of exposure, and vulnerabilities of elements at risk. Reducing the severity of each of these components is the objective of flood risk management under the UNISDR guideline and idea of “living with floods”. On the basis of this framework, the present research reviews flood risk within the Lagos area of Nigeria over the period 1968–2012. During this period, floods have caused harm to millions of people physically, emotionally, and economically. Arguably over this period the efforts of stakeholders to address the challenges appear to have been limited by, amongst other things, a lack of reliable data, a lack of awareness amongst the population affected, and a lack of knowledge of flood risk mitigation. It is the aim of this research to assess the current understanding of flood risk and management in Lagos and to offer recommendations towards future guidance.

scholarly journals Multi-variate flood damage assessment: a tree-based data-mining approach

2013 ◽  
Vol 13 (1) ◽  
pp. 53-64 ◽  
Author(s):  
B. Merz ◽  
H. Kreibich ◽  
U. Lall
Keyword(s):  
Data Mining ◽  
Damage Assessment ◽  
Economic Status ◽  
Flood Damage ◽  
Damage Functions ◽  
Data Set ◽  
Promising Alternative ◽  
Damage Models ◽  
Inundation Depth ◽  
Flood Damage Assessment

Abstract. The usual approach for flood damage assessment consists of stage-damage functions which relate the relative or absolute damage for a certain class of objects to the inundation depth. Other characteristics of the flooding situation and of the flooded object are rarely taken into account, although flood damage is influenced by a variety of factors. We apply a group of data-mining techniques, known as tree-structured models, to flood damage assessment. A very comprehensive data set of more than 1000 records of direct building damage of private households in Germany is used. Each record contains details about a large variety of potential damage-influencing characteristics, such as hydrological and hydraulic aspects of the flooding situation, early warning and emergency measures undertaken, state of precaution of the household, building characteristics and socio-economic status of the household. Regression trees and bagging decision trees are used to select the more important damage-influencing variables and to derive multi-variate flood damage models. It is shown that these models outperform existing models, and that tree-structured models are a promising alternative to traditional damage models.

Sign in / Sign up

Export Citation Format

Share Document