European Member States are required to conduct a comprehensive flood risk assessment that accounts for people, the environment, cultural heritage, and economic activities under the European Floods Directive (2007/60/EC). Such an assessment is considered key knowledge for the development of efficient flood risk management plans. Although limiting the damage to the population is a primary objective of flood risk management, models for assessing damage to people are scarce. Current assessment practices often adopt the number of potentially affected individuals, specifically those living in the flooded area, as a proxy for assessing damage to people. While practical, such an approach ignores the broad flood consequences on population well-being. Empirical studies have shown that the flood negative consequences on people encompass a wide range of direct and indirect, tangible and intangible impact types. Some of these impacts may extend over time, resulting in long-term consequences, and across space, affecting individuals who live beyond the flooded area. The lack of models capable of capturing these impact mechanisms and accounting for all segments of society affected by them reflects the absence of a comprehensive understanding of the overall impact of floods on people. This PhD research aims to address these gaps by investigating the "overall impact" of floods on people, defined as the subjective perception of the broad range of flood consequences experienced by individuals. Specifically, the research aims to conceptualize the overall impact across differently exposed groups of society and to develop a predictive model for spatial application, ultimately providing insights into how overall impact on people can be reduced. The analyses are grounded in the case study of the Marche region in Italy, which was affected by an exceptional flood event in September 2022. The analysis draws on original data collected in the affected area through an initial field survey and a subsequent online questionnaire. This fieldwork not only facilitated the creation of a post-event physical damage dataset but also enabled the collection of spontaneous narratives from affected residents. The insights gained in the field informed the design of the online questionnaire, which was administered to residents of the affected communities 20 months after the event to investigate the overall impact of the flood on people. The questionnaire targeted directly affected individuals (i.e., those whose homes were flooded), indirectly affected individuals (e.g., those experiencing work interruption), and not affected respondents, and collected information on hazard characteristics, vulnerability variables, and a wide range of flood impacts. These included both direct and indirect, tangible and intangible impacts, measured through perceived severity scales and objective metrics. Data were collected from 707 respondents. Descriptive statistics are employed to explore how the perceived severity of the different impact types varies across the three exposed groups. Regression analyses are conducted separately for each exposed group to assess the role of the different perceived impact types in explaining variations in overall flood impact. The findings of these analyses clarify the mechanisms through which floods affect people, highlighting the role of indirect and intangible flood consequences. The analyses further show that objectively measurable impact variables explain only a limited share of overall impact compared with perceived impact variables, indicating that appraisal-based data remains necessary to measure the broader consequences of floods on people wellbeing. Regression models are used to examine the extent to which overall impact can be explained by exposure conditions, hazard intensity, and vulnerability variables, thereby supporting the development of a predictive model. The results show that these variables explain 38% of the variance in overall impact, with exposure conditions accounting for the largest contribution. Among vulnerability variables, only a limited subset of variable shows significant effects, suggesting that socio-economic and demographic variables alone do not adequately represent the individuals’ vulnerability to overall flood impact. The developed predictive model is applied to the Areas of Potential Significant Flood Risk (APSFRs) of the Parma and Baganza rivers. Such an application shows how the proposed approach can support flood damage assessment beyond current practice by estimating the expected intensity of impact on people rather than relying solely on counts of affected residents. Finally, the findings are discussed from both theoretical and practical perspectives in order to inform future research and support flood risk management.
La Direttiva Europea sulle Alluvioni (2007/60/CE) richiede agli Stati membri di effettuare una valutazione comprensiva del rischio alluvionale che consideri persone, ambiente, patrimonio culturale e attività economiche. Tale valutazione costituisce una base conoscitiva fondamentale per la definizione di piani gestione del rischio efficienti. Tuttavia, nonostante la riduzione dei danni alla popolazione costituisca l’obiettivo principale della gestione del rischio alluvionale, i modelli disponibili per stimare tali danni sono ancora limitati. In particolare, la pratica corrente si basa su valutazioni semplicistiche del danno, utilizzando la proxy del numero di persone potenzialmente colpite dall’evento, ovvero il numero di residenti che risiedono nell’area allagata. Sebbene questo approccio sia di semplice applicazione, esso non è in grado di rappresentare l’ampio spettro di conseguenze che le alluvioni possono causare al benessere delle persone. La letteratura empirica evidenzia come gli impatti delle alluvioni includano conseguenze dirette e indirette, tangibili e intangibili, che possono persistere nel tempo ed estendersi oltre le aree allagate, interessando anche individui non direttamente esposti all’evento. L’assenza di modelli capaci di rappresentare tali meccanismi di danno per tutti i segmenti della società che possono essere colpiti da un evento riflette una comprensione ancora limitata dell’impatto complessivo delle alluvioni sulla popolazione. Questa ricerca di dottorato si propone di colmare tali lacune attraverso lo studio del “danno complessivo” delle alluvioni sulle persone, definito come la percezione soggettiva dell’insieme delle conseguenze sperimentate dagli individui a causa dell’evento. In particolare, gli obiettivi della ricerca sono la concettualizzazione del danno complessivo nelle diverse categorie di popolazione esposta e lo sviluppo di un modello predittivo applicabile alla scala della sezione di censimento, al fine di fornire indicazioni utili alla riduzione degli impatti delle alluvioni sulla popolazione. Le analisi si basano sul caso studio della regione Marche, la quale è stata colpita da evento alluvionale di eccezionale intensità a settembre 2022. La ricerca utilizza dati raccolti attraverso una campagna di rilievo sul campo e un questionario online. Le attività sul campo hanno consentito la costruzione di un database dei danni fisici post-evento e la raccolta di testimonianze dei residenti colpiti. Tali testimonianze hanno guidato la progettazione del questionario, somministrato venti mesi dopo l’evento ai residenti delle comunità interessate. Il questionario è stato rivolto a persone direttamente colpite dall’alluvione (ovvero residenti le cui abitazioni sono state allagate), persone indirettamente colpite (ad esempio coloro che hanno dovuto interrompere l’attività lavorativa) e soggetti non colpiti né direttamente né indirettamente. Il questionario è stato progettato per raccogliere informazioni relative alle caratteristiche dell’evento, alle caratteristiche di vulnerabilità della popolazione e a diversi tipi di impatti da alluvione. Tali impatti comprendono conseguenze dirette e indirette, tangibili e intangibili, misurate sia attraverso scale di severità percepita sia mediante indicatori oggettivi. Complessivamente sono state raccolte 707 risposte. Analisi statistiche descrittive sono state inizialmente utilizzate per analizzare le differenze nella severità percepita delle diverse tipologie di impatto tra i gruppi di esposizione. Successivamente, modelli di regressione separati per ciascun gruppo sono stati impiegati per valutare il ruolo delle diverse tipologie di impatto nella determinazione del danno complessivo percepito. I risultati chiariscono i meccanismi attraverso cui le alluvioni incidono sul benessere delle persone, evidenziando l’importanza delle conseguenze indirette e intangibili. Le analisi mostrano inoltre che le variabili di impatto oggettive spiegano solo una parte limitata del danno complessivo rispetto alle variabili percettive, sottolineando il ruolo centrale delle variabili di natura soggettiva nel cogliere le conseguenze più ampie delle alluvioni sul benessere della popolazione. Ulteriori modelli di regressione sono stati impiegati per analizzare in quale misura il danno complessivo possa essere spiegato dal grado di esposizione, dall’intensità dell’evento e dalle caratteristiche di vulnerabilità individuale. I risultati mostrano che tali variabili spiegano il 38% della variabilità del danno complessivo. Tra le variabili esplicative, il grado di esposizione rappresenta il contributo più rilevante. Tra le variabili di vulnerabilità, solo un numero limitato risulta essere significativo, suggerendo che le sole caratteristiche socioeconomiche e demografiche non siano sufficienti a rappresentare adeguatamente la vulnerabilità degli individui al danno complessivo delle alluvioni. Il modello sviluppato è stato applicato alle Aree a Potenziale Rischio Significativo di Alluvione (APSFR) dei fiumi Parma e Baganza. L’applicazione mostra come l’approccio proposto consenta di migliorare la valutazione dei danni alla popolazione rispetto alle pratiche correnti, stimando l’intensità attesa del danno anziché il numero di residenti esposti. I risultati raggiunti sono discussi sia dal punto di vista teorico sia applicativo, al fine di contribuire all’avanzamento della ricerca sugli impatti delle alluvioni e di supportare lo sviluppo di strategie più efficaci per la gestione del rischio alluvionale.
Beyond visible damage: conceptualising and modelling flood impacts on people using survey data
RROKAJ, SARA
2025/2026
Abstract
European Member States are required to conduct a comprehensive flood risk assessment that accounts for people, the environment, cultural heritage, and economic activities under the European Floods Directive (2007/60/EC). Such an assessment is considered key knowledge for the development of efficient flood risk management plans. Although limiting the damage to the population is a primary objective of flood risk management, models for assessing damage to people are scarce. Current assessment practices often adopt the number of potentially affected individuals, specifically those living in the flooded area, as a proxy for assessing damage to people. While practical, such an approach ignores the broad flood consequences on population well-being. Empirical studies have shown that the flood negative consequences on people encompass a wide range of direct and indirect, tangible and intangible impact types. Some of these impacts may extend over time, resulting in long-term consequences, and across space, affecting individuals who live beyond the flooded area. The lack of models capable of capturing these impact mechanisms and accounting for all segments of society affected by them reflects the absence of a comprehensive understanding of the overall impact of floods on people. This PhD research aims to address these gaps by investigating the "overall impact" of floods on people, defined as the subjective perception of the broad range of flood consequences experienced by individuals. Specifically, the research aims to conceptualize the overall impact across differently exposed groups of society and to develop a predictive model for spatial application, ultimately providing insights into how overall impact on people can be reduced. The analyses are grounded in the case study of the Marche region in Italy, which was affected by an exceptional flood event in September 2022. The analysis draws on original data collected in the affected area through an initial field survey and a subsequent online questionnaire. This fieldwork not only facilitated the creation of a post-event physical damage dataset but also enabled the collection of spontaneous narratives from affected residents. The insights gained in the field informed the design of the online questionnaire, which was administered to residents of the affected communities 20 months after the event to investigate the overall impact of the flood on people. The questionnaire targeted directly affected individuals (i.e., those whose homes were flooded), indirectly affected individuals (e.g., those experiencing work interruption), and not affected respondents, and collected information on hazard characteristics, vulnerability variables, and a wide range of flood impacts. These included both direct and indirect, tangible and intangible impacts, measured through perceived severity scales and objective metrics. Data were collected from 707 respondents. Descriptive statistics are employed to explore how the perceived severity of the different impact types varies across the three exposed groups. Regression analyses are conducted separately for each exposed group to assess the role of the different perceived impact types in explaining variations in overall flood impact. The findings of these analyses clarify the mechanisms through which floods affect people, highlighting the role of indirect and intangible flood consequences. The analyses further show that objectively measurable impact variables explain only a limited share of overall impact compared with perceived impact variables, indicating that appraisal-based data remains necessary to measure the broader consequences of floods on people wellbeing. Regression models are used to examine the extent to which overall impact can be explained by exposure conditions, hazard intensity, and vulnerability variables, thereby supporting the development of a predictive model. The results show that these variables explain 38% of the variance in overall impact, with exposure conditions accounting for the largest contribution. Among vulnerability variables, only a limited subset of variable shows significant effects, suggesting that socio-economic and demographic variables alone do not adequately represent the individuals’ vulnerability to overall flood impact. The developed predictive model is applied to the Areas of Potential Significant Flood Risk (APSFRs) of the Parma and Baganza rivers. Such an application shows how the proposed approach can support flood damage assessment beyond current practice by estimating the expected intensity of impact on people rather than relying solely on counts of affected residents. Finally, the findings are discussed from both theoretical and practical perspectives in order to inform future research and support flood risk management.| File | Dimensione | Formato | |
|---|---|---|---|
|
dissertation_Rrokaj.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: tesi di dottorato
Dimensione
6.78 MB
Formato
Adobe PDF
|
6.78 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/259240