@phdthesis{Schoppa2023,
  author    = {Schoppa, Lukas},
  title     = {Dynamics in the flood vulnerability of companies},
  doi       = {10.25932/publishup-59242},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-592424},
  school      = {Universit{\"a}t Potsdam},
  pages     = {X, 165},
  year      = {2023},
  abstract  = {River flooding is a constant peril for societies, causing direct economic losses in the order of \$100 billion worldwide each year. Under global change, the prolonged concentration of people and assets in floodplains is accompanied by an emerging intensification of flood extremes due to anthropogenic global warming, ultimately exacerbating flood risk in many regions of the world. Flood adaptation plays a key role in the mitigation of impacts, but poor understanding of vulnerability and its dynamics limits the validity of predominant risk assessment methods and impedes effective adaptation strategies. Therefore, this thesis investigates new methods for flood risk assessment that embrace the complexity of flood vulnerability, using the understudied commercial sector as an application example. Despite its importance for accurate risk evaluation, flood loss modeling has been based on univariable and deterministic stage-damage functions for a long time. However, such simplistic methods only insufficiently describe the large variation in damage processes, which initiated the development of multivariable and probabilistic loss estimation techniques. The first study of this thesis developed flood loss models for companies that are based on emerging statistical and machine learning approaches (i.e., random forest, Bayesian network, Bayesian regression). In a benchmarking experiment on basis of object-level loss survey data, the study showed that all proposed models reproduced the heterogeneity in damage processes and outperformed conventional stage-damage functions with respect to predictive accuracy. Another advantage of the novel methods is that they convey probabilistic information in predictions, which communicates the large remaining uncertainties transparently and, hence, supports well-informed risk assessment. Flood risk assessment combines vulnerability assessment (e.g., loss estimation) with hazard and exposure analyses. Although all of the three risk drivers interact and change over time, such dependencies and dynamics are usually not explicitly included in flood risk models. Recently, systemic risk assessment that dissolves the isolated consideration of risk drivers has gained traction, but the move to holistic risk assessment comes with limited thoroughness in terms of loss estimation and data limitations. In the second study, I augmented a socio-hydrological system dynamics model for companies in Dresden, Germany, with the multivariable Bayesian regression loss model from the first study. The additional process-detail and calibration data improved the loss estimation in the systemic risk assessment framework and contributed to more accurate and reliable simulations. The model uses Bayesian inference to quantify uncertainty and learn the model parameters from a combination of prior knowledge and diverse data. The third study demonstrates the potential of the socio-hydrological flood risk model for continuous, long-term risk assessment and management. Using hydroclimatic ad socioeconomic forcing data, I projected a wide range of possible risk trajectories until the end of the century, taking into account the adaptive behavior of companies. The study results underline the necessity of increased adaptation efforts to counteract the expected intensification of flood risk due to climate change. A sensitivity analysis of the effectiveness of different adaptation measures and strategies revealed that optimized adaptation has the potential to mitigate flood risk by up to 60\%, particularly when combining structural and non-structural measures. Additionally, the application shows that systemic risk assessment is capable of capturing adverse long-term feedbacks in the human-flood system such as the levee effect. Overall, this thesis advances the representation of vulnerability in flood risk modeling by offering modeling solutions that embrace the complexity of human-flood interactions and quantify uncertainties consistently using probabilistic modeling. The studies show how scarce information in data and previous experiments can be integrated in the inference process to provide model predictions and simulations that are reliable and rich in information. Finally, the focus on the flood vulnerability of companies provides new insights into the heterogeneous damage processes and distinct flood coping of this sector.},
  language  = {en}
}
@phdthesis{Brill2022,
  author    = {Brill, Fabio Alexander},
  title     = {Applications of machine learning and open geospatial data in flood risk modelling},
  doi       = {10.25932/publishup-55594},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-555943},
  school      = {Universit{\"a}t Potsdam},
  pages     = {xix, 124},
  year      = {2022},
  abstract  = {Der technologische Fortschritt erlaubt es, zunehmend komplexe Vorhersagemodelle auf Basis immer gr{\"o}ßerer Datens{\"a}tze zu produzieren. F{\"u}r das Risikomanagement von Naturgefahren sind eine Vielzahl von Modellen als Entscheidungsgrundlage notwendig, z.B. in der Auswertung von Beobachtungsdaten, f{\"u}r die Vorhersage von Gefahrenszenarien, oder zur statistischen Absch{\"a}tzung der zu erwartenden Sch{\"a}den. Es stellt sich also die Frage, inwiefern moderne Modellierungsans{\"a}tze wie das maschinelle Lernen oder Data-Mining in diesem Themenbereich sinnvoll eingesetzt werden k{\"o}nnen. Zus{\"a}tzlich ist im Hinblick auf die Datenverf{\"u}gbarkeit und -zug{\"a}nglichkeit ein Trend zur {\"O}ffnung (open data) zu beobachten. Thema dieser Arbeit ist daher, die M{\"o}glichkeiten und Grenzen des maschinellen Lernens und frei verf{\"u}gbarer Geodaten auf dem Gebiet der Hochwasserrisikomodellierung im weiteren Sinne zu untersuchen. Da dieses {\"u}bergeordnete Thema sehr breit ist, werden einzelne relevante Aspekte herausgearbeitet und detailliert betrachtet. Eine prominente Datenquelle im Bereich Hochwasser ist die satellitenbasierte Kartierung von {\"U}berflutungsfl{\"a}chen, die z.B. {\"u}ber den Copernicus Service der Europ{\"a}ischen Union frei zur Verf{\"u}gung gestellt werden. Große Hoffnungen werden in der wissenschaftlichen Literatur in diese Produkte gesetzt, sowohl f{\"u}r die akute Unterst{\"u}tzung der Einsatzkr{\"a}fte im Katastrophenfall, als auch in der Modellierung mittels hydrodynamischer Modelle oder zur Schadensabsch{\"a}tzung. Daher wurde ein Fokus in dieser Arbeit auf die Untersuchung dieser Flutmasken gelegt. Aus der Beobachtung, dass die Qualit{\"a}t dieser Produkte in bewaldeten und urbanen Gebieten unzureichend ist, wurde ein Verfahren zur nachtr{\"a}glichenVerbesserung mittels maschinellem Lernen entwickelt. Das Verfahren basiert auf einem Klassifikationsalgorithmus der nur Trainingsdaten von einer vorherzusagenden Klasse ben{\"o}tigt, im konkreten Fall also Daten von {\"U}berflutungsfl{\"a}chen, nicht jedoch von der negativen Klasse (trockene Gebiete). Die Anwendung f{\"u}r Hurricane Harvey in Houston zeigt großes Potenzial der Methode, abh{\"a}ngig von der Qualit{\"a}t der urspr{\"u}nglichen Flutmaske. Anschließend wird anhand einer prozessbasierten Modellkette untersucht, welchen Einfluss implementierte physikalische Prozessdetails auf das vorhergesagte statistische Risiko haben. Es wird anschaulich gezeigt, was eine Risikostudie basierend auf etablierten Modellen leisten kann. Solche Modellketten sind allerdings bereits f{\"u}r Flusshochwasser sehr komplex, und f{\"u}r zusammengesetzte oder kaskadierende Ereignisse mit Starkregen, Sturzfluten, und weiteren Prozessen, kaum vorhanden. Im vierten Kapitel dieser Arbeit wird daher getestet, ob maschinelles Lernen auf Basis von vollst{\"a}ndigen Schadensdaten einen direkteren Weg zur Schadensmodellierung erm{\"o}glicht, der die explizite Konzeption einer solchen Modellkette umgeht. Dazu wird ein staatlich erhobener Datensatz der gesch{\"a}digten Geb{\"a}ude w{\"a}hrend des schweren El Ni{\~n}o Ereignisses 2017 in Peru verwendet. In diesem Kontext werden auch die M{\"o}glichkeiten des Data-Mining zur Extraktion von Prozessverst{\"a}ndnis ausgelotet. Es kann gezeigt werden, dass diverse frei verf{\"u}gbare Geodaten n{\"u}tzliche Informationen f{\"u}r die Gefahren- und Schadensmodellierung von komplexen Flutereignissen liefern, z.B. satellitenbasierte Regenmessungen, topographische und hydrographische Information, kartierte Siedlungsfl{\"a}chen, sowie Indikatoren aus Spektraldaten. Zudem zeigen sich Erkenntnisse zu den Sch{\"a}digungsprozessen, die im Wesentlichen mit den vorherigen Erwartungen in Einklang stehen. Die maximale Regenintensit{\"a}t wirkt beispielsweise in St{\"a}dten und steilen Schluchten st{\"a}rker sch{\"a}digend, w{\"a}hrend die Niederschlagssumme in tiefliegenden Flussgebieten und bewaldeten Regionen als aussagekr{\"a}ftiger befunden wurde. L{\"a}ndliche Gebiete in Peru weisen in der pr{\"a}sentierten Studie eine h{\"o}here Vulnerabilit{\"a}t als die Stadtgebiete auf. Jedoch werden auch die grunds{\"a}tzlichen Grenzen der Methodik und die Abh{\"a}ngigkeit von spezifischen Datens{\"a}tzen and Algorithmen offenkundig. In der {\"u}bergreifenden Diskussion werden schließlich die verschiedenen Methoden - prozessbasierte Modellierung, pr{\"a}diktives maschinelles Lernen, und Data-Mining - mit Blick auf die Gesamtfragestellungen evaluiert. Im Bereich der Gefahrenbeobachtung scheint eine Fokussierung auf neue Algorithmen sinnvoll. Im Bereich der Gefahrenmodellierung, insbesondere f{\"u}r Flusshochwasser, wird eher die Verbesserung von physikalischen Modellen, oder die Integration von prozessbasierten und statistischen Verfahren angeraten. In der Schadensmodellierung fehlen nach wie vor die großen repr{\"a}sentativen Datens{\"a}tze, die f{\"u}r eine breite Anwendung von maschinellem Lernen Voraussetzung ist. Daher ist die Verbesserung der Datengrundlage im Bereich der Sch{\"a}den derzeit als wichtiger einzustufen als die Auswahl der Algorithmen.},
  language  = {en}
}
@phdthesis{Sieg2018,
  author    = {Sieg, Tobias},
  title     = {Reliability of flood damage estimations across spatial scales},
  doi       = {10.25932/publishup-42616},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-426161},
  school      = {Universit{\"a}t Potsdam},
  pages     = {XIII, 115},
  year      = {2018},
  abstract  = {Extreme Naturereignisse sind ein integraler Bestandteil der Natur der Erde. Sie werden erst dann zu Gefahren f{\"u}r die Gesellschaft, wenn sie diesen Ereignissen ausgesetzt ist. Dann allerdings k{\"o}nnen Naturgefahren verheerende Folgen f{\"u}r die Gesellschaft haben. Besonders hydro-meteorologische Gefahren wie zum Beispiel Flusshochwasser, Starkregenereignisse, Winterst{\"u}rme, Orkane oder Tornados haben ein hohes Schadenspotential und treten rund um den Globus auf. Einhergehend mit einer immer w{\"a}rmer werdenden Welt, werden auch Extremwetterereignisse, welche potentiell Naturgefahren ausl{\"o}sen k{\"o}nnen, immer wahrscheinlicher. Allerdings tr{\"a}gt nicht nur eine sich ver{\"a}ndernde Umwelt zur Erh{\"o}hung des Risikos von Naturgefahren bei, sondern auch eine sich ver{\"a}ndernde Gesellschaft. Daher ist ein angemessenes Risikomanagement erforderlich um die Gesellschaft auf jeder r{\"a}umlichen Ebene an diese Ver{\"a}nderungen anzupassen. Ein essentieller Bestandteil dieses Managements ist die Absch{\"a}tzung der {\"o}konomischen Auswirkungen der Naturgefahren. Bisher allerdings fehlen verl{\"a}ssliche Methoden um die Auswirkungen von hydro-meteorologischen Gefahren abzusch{\"a}tzen. Ein Hauptbestandteil dieser Arbeit ist daher die Entwicklung und Anwendung einer neuen Methode, welche die Verl{\"a}sslichkeit der Schadenssch{\"a}tzung verbessert. Die Methode wurde beispielhaft zur Sch{\"a}tzung der {\"o}konomischen Auswirkungen eines Flusshochwassers auf einzelne Unternehmen bis hin zu den Auswirkungen auf das gesamte Wirtschaftssystem Deutschlands erfolgreich angewendet. Bestehende Methoden geben meist wenig Information {\"u}ber die Verl{\"a}sslichkeit ihrer Sch{\"a}tzungen. Da diese Informationen Entscheidungen zur Anpassung an das Risiko erleichtern, wird die Verl{\"a}sslichkeit der Schadenssch{\"a}tzungen mit der neuen Methode dargestellt. Die Verl{\"a}sslichkeit bezieht sich dabei nicht nur auf die Schadenssch{\"a}tzung selber, sondern auch auf die Annahmen, die {\"u}ber betroffene Geb{\"a}ude gemacht werden. Nach diesem Prinzip kann auch die Verl{\"a}sslichkeit von Annahmen {\"u}ber die Zukunft dargestellt werden, dies ist ein wesentlicher Aspekt f{\"u}r Prognosen. Die Darstellung der Verl{\"a}sslichkeit und die erfolgreiche Anwendung zeigt das Potential der Methode zur Verwendung von Analysen f{\"u}r gegenw{\"a}rtige und zuk{\"u}nftige hydro-meteorologische Gefahren.},
  language  = {en}
}
@phdthesis{Roezer2018,
  author    = {R{\"o}zer, Viktor},
  title     = {Pluvial flood loss to private households},
  doi       = {10.25932/publishup-42991},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-429910},
  school      = {Universit{\"a}t Potsdam},
  pages     = {XXII, 109},
  year      = {2018},
  abstract  = {Today, more than half of the world's population lives in urban areas. With a high density of population and assets, urban areas are not only the economic, cultural and social hubs of every society, they are also highly susceptible to natural disasters. As a consequence of rising sea levels and an expected increase in extreme weather events caused by a changing climate in combination with growing cities, flooding is an increasing threat to many urban agglomerations around the globe. To mitigate the destructive consequences of flooding, appropriate risk management and adaptation strategies are required. So far, flood risk management in urban areas is almost exclusively focused on managing river and coastal flooding. Often overlooked is the risk from small-scale rainfall-triggered flooding, where the rainfall intensity of rainstorms exceeds the capacity of urban drainage systems, leading to immediate flooding. Referred to as pluvial flooding, this flood type exclusive to urban areas has caused severe losses in cities around the world. Without further intervention, losses from pluvial flooding are expected to increase in many urban areas due to an increase of impervious surfaces compounded with an aging drainage infrastructure and a projected increase in heavy precipitation events. While this requires the integration of pluvial flood risk into risk management plans, so far little is known about the adverse consequences of pluvial flooding due to a lack of both detailed data sets and studies on pluvial flood impacts. As a consequence, methods for reliably estimating pluvial flood losses, needed for pluvial flood risk assessment, are still missing. Therefore, this thesis investigates how pluvial flood losses to private households can be reliably estimated, based on an improved understanding of the drivers of pluvial flood loss. For this purpose, detailed data from pluvial flood-affected households was collected through structured telephone- and web-surveys following pluvial flood events in Germany and the Netherlands. Pluvial flood losses to households are the result of complex interactions between impact characteristics such as the water depth and a household's resistance as determined by its risk awareness, preparedness, emergency response, building properties and other influencing factors. Both exploratory analysis and machine-learning approaches were used to analyze differences in resistance and impacts between households and their effects on the resulting losses. The comparison of case studies showed that the awareness around pluvial flooding among private households is quite low. Low awareness not only challenges the effective dissemination of early warnings, but was also found to influence the implementation of private precautionary measures. The latter were predominately implemented by households with previous experience of pluvial flooding. Even cases where previous flood events affected a different part of the same city did not lead to an increase in preparedness of the surveyed households, highlighting the need to account for small-scale variability in both impact and resistance parameters when assessing pluvial flood risk. While it was concluded that the combination of low awareness, ineffective early warning and the fact that only a minority of buildings were adapted to pluvial flooding impaired the coping capacities of private households, the often low water levels still enabled households to mitigate or even prevent losses through a timely and effective emergency response. These findings were confirmed by the detection of loss-influencing variables, showing that cases in which households were able to prevent any loss to the building structure are predominately explained by resistance variables such as the household's risk awareness, while the degree of loss is mainly explained by impact variables. Based on the important loss-influencing variables detected, different flood loss models were developed. Similar to flood loss models for river floods, the empirical data from the preceding data collection was used to train flood loss models describing the relationship between impact and resistance parameters and the resulting loss to building structures. Different approaches were adapted from river flood loss models using both models with the water depth as only predictor for building structure loss and models incorporating additional variables from the preceding variable detection routine. The high predictive errors of all compared models showed that point predictions are not suitable for estimating losses on the building level, as they severely impair the reliability of the estimates. For that reason, a new probabilistic framework based on Bayesian inference was introduced that is able to provide predictive distributions instead of single loss estimates. These distributions not only give a range of probable losses, they also provide information on how likely a specific loss value is, representing the uncertainty in the loss estimate. Using probabilistic loss models, it was found that the certainty and reliability of a loss estimate on the building level is not only determined by the use of additional predictors as shown in previous studies, but also by the choice of response distribution defining the shape of the predictive distribution. Here, a mix between a beta and a Bernoulli distribution to account for households that are able to prevent losses to their building's structure was found to provide significantly more certain and reliable estimates than previous approaches using Gaussian or non-parametric response distributions. The successful model transfer and post-event application to estimate building structure loss in Houston, TX, caused by pluvial flooding during Hurricane Harvey confirmed previous findings, and demonstrated the potential of the newly developed multi-variable beta model for future risk assessments. The highly detailed input data set constructed from openly available data sources containing over 304,000 affected buildings in Harris County further showed the potential of data-driven, building-level loss models for pluvial flood risk assessment. In conclusion, pluvial flood losses to private households are the result of complex interactions between impact and resistance variables, which should be represented in loss models. The local occurrence of pluvial floods requires loss estimates on high spatial resolutions, i.e. on the building level, where losses are variable and uncertainties are high. Therefore, probabilistic loss estimates describing the uncertainty of the estimate should be used instead of point predictions. While the performance of probabilistic models on the building level are mainly driven by the choice of response distribution, multi-variable models are recommended for two reasons: First, additional resistance variables improve the detection of cases in which households were able to prevent structural losses. Second, the added variability of additional predictors provides a better representation of the uncertainties when loss estimates from multiple buildings are aggregated. This leads to the conclusion that data-driven probabilistic loss models on the building level allow for a reliable loss estimation at an unprecedented level of detail, with a consistent quantification of uncertainties on all aggregation levels. This makes the presented approach suitable for a wide range of applications, from decision support in spatial planning to impact- based early warning systems.},
  language  = {en}
}
@phdthesis{Kellermann2017,
  author    = {Kellermann, Patric},
  title     = {Assessing natural risks for railway infrastructure and transportation in Austria},
  url       = {http://nbn-resolving.de/urn:nbn:de:kobv:517-opus4-103877},
  school      = {Universit{\"a}t Potsdam},
  pages     = {x, 113},
  year      = {2017},
  abstract  = {Natural hazards can have serious societal and economic impacts. Worldwide, around one third of economic losses due to natural hazards are attributable to floods. The majority of natural hazards are triggered by weather-related extremes such as heavy precipitation, rapid snow melt, or extreme temperatures. Some of them, and in particular floods, are expected to further increase in terms of frequency and/or intensity in the coming decades due to the impacts of climate change. In this context, the European Alps areas are constantly disclosed as being particularly sensitive. In order to enhance the resilience of societies to natural hazards, risk assessments are substantial as they can deliver comprehensive risk information to be used as a basis for effective and sustainable decision-making in natural hazards management. So far, current assessment approaches mostly focus on single societal or economic sectors - e.g. flood damage models largely concentrate on private-sector housing - and other important sectors, such as the transport infrastructure sector, are widely neglected. However, transport infrastructure considerably contributes to economic and societal welfare, e.g. by ensuring mobility of people and goods. In Austria, for example, the national railway network is essential for the European transit of passengers and freights as well as for the development of the complex Alpine topography. Moreover, a number of recent experiences show that railway infrastructure and transportation is highly vulnerable to natural hazards. As a consequence, the Austrian Federal Railways had to cope with economic losses on the scale of several million euros as a result of flooding and other alpine hazards. The motivation of this thesis is to contribute to filling the gap of knowledge about damage to railway infrastructure caused by natural hazards by providing new risk information for actors and stakeholders involved in the risk management of railway transportation. Hence, in order to support the decision-making towards a more effective and sustainable risk management, the following two shortcomings in natural risks research are approached: i) the lack of dedicated models to estimate flood damage to railway infrastructure, and ii) the scarcity of insights into possible climate change impacts on the frequency of extreme weather events with focus on future implications for railway transportation in Austria. With regard to flood impacts to railway infrastructure, the empirically derived damage model Railway Infrastructure Loss (RAIL) proved expedient to reliably estimate both structural flood damage at exposed track sections of the Northern Railway and resulting repair cost. The results show that the RAIL model is capable of identifying flood risk hot spots along the railway network and, thus, facilitates the targeted planning and implementation of (technical) risk reduction measures. However, the findings of this study also show that the development and validation of flood damage models for railway infrastructure is generally constrained by the continuing lack of detailed event and damage data. In order to provide flood risk information on the large scale to support strategic flood risk management, the RAIL model was applied for the Austrian Mur River catchment using three different hydraulic scenarios as input as well as considering an increased risk aversion of the railway operator. Results indicate that the model is able to deliver comprehensive risk information also on the catchment level. It is furthermore demonstrated that the aspect of risk aversion can have marked influence on flood damage estimates for the study area and, hence, should be considered with regard to the development of risk management strategies. Looking at the results of the investigation on future frequencies of extreme weather events jeopardizing railway infrastructure and transportation in Austria, it appears that an increase in intense rainfall events and heat waves has to be expected, whereas heavy snowfall and cold days are likely to decrease. Furthermore, results indicate that frequencies of extremes are rather sensitive to changes of the underlying thresholds. It thus emphasizes the importance to carefully define, validate, and — if needed — to adapt the thresholds that are used to detect and forecast meteorological extremes. For this, continuous and standardized documentation of damaging events and near-misses is a prerequisite. Overall, the findings of the research presented in this thesis agree on the necessity to improve event and damage documentation procedures in order to enable the acquisition of comprehensive and reliable risk information via risk assessments and, thus, support strategic natural hazards management of railway infrastructure and transportation.},
  language  = {en}
}