Veranstaltungsprogramm

Jeju Island, a volcanic island in Korea, exhibits distinct hydrogeomorphological features that pose challenges for disaster management. Traditional hydrological models and qualitative risk assessment methods developed for inland areas are insufficient for addressing Jeju Island’s unique conditions. A localized approach is necessary to integrate hydrogeomorphological and urban drainage factors with hazard-related data. This study develops a grid-based AI flood risk assessment method tailored to Jeju Island’s characteristics, considering urban drainage systems to improve prediction accuracy. The objectives are to: (1) establish a grid-type flood damage and flood risk influencing factor database that reflects Jeju Island's unique hydrogeomorphological and hazard characteristics, (2) create a binary classification deep learning model integrating hydrological, hazard, and urban drainage factors to assess flood risk, and (3) propose warning standards based on flood damage data and risk assessments, accounting for the specific disaster patterns of Jeju Island. This study seeks to overcome the limitations of existing disaster management approaches by leveraging AI and hydrological analysis technologies, addressing Jeju Island's increasing vulnerability to disasters. Through the integration of grid-based data and advanced ML models, this research aims to provide a comprehensive framework for predicting and managing flood risks, ultimately enhancing resilience and safety on the area.

Any disruptions in urban stormwater networks (USNs), such as sewer or manhole failures, can propagate impacts beyond their immediate spatial location (e.g., upstream or downstream). However, determining where and to what extent these impacts propagate is often challenging due to the complex hydraulic behaviour of such systems. This study introduces a modelling framework using graph theory to represent the consequences of failure propagations. The framework quantifies the impacts of junction failures on other elements of USNs through hydrodynamic modelling. These impacts are then decoupled from the physical USN topology and represented as a directed and weighted graph, termed "impact network", where edge weights capture the actual consequences beyond the failure’s point of origin. The application of the proposed framework was demonstrated through systematic investigations of a real-world case study. This analysis led to the identification of the most influential nodes driving impact propagation and the most affected nodes exposed to the propagated impacts. Furthermore, the spatial interconnections between the impact network and critical urban locations were analysed, enabling the prioritisation of nodes of strategic importance.

In the Ruhr region of Germany, pumping stations are often used in urban areas to drain water from rivers due to land subsidence caused by coal mining. In case of pump failure, these areas face an immense threat to human life and property due to the short warning time of flooding. Data-driven models can predict flood inundation in real time, allowing for the timely initiation of protective measures. In this study, a new approach based on a data-driven model is developed using a convolutional neural network (CNN) to dynamically predict the maximum water depth of the next 24 h for fluvial and pluvial flooding in real time. The model is trained with physically based pre-simulated scenarios considering rainfall and runoff curves as inputs. The resulting forecasting system provides accurate real-time predictions of flood extent and water depth for different pump failure scenarios up to an average root mean square error of 0.038 m and an average critical success index of 0.953. The developed forecasting system is a suitable approach for operational control systems that allow accurate prediction in real time. A further research issue represents the dynamic forecast by the developed approach, predicting the current water depth over an event.

With climate change and urbanization, urban flooding has intensified, causing significant socio-economic damages. Existing studies lack evaluation of flooding risks from a national scale or multi-city comparative perspective. This study aims to assess the urban flooding risks of 332 prefecture-level cities in China under the Shared Socioeconomic Pathway-Representative Concentration Pathway (SSP-RCP) SSP2-RCP4.5 and SSP5-RCP8.5 for the years 2020, 2040, 2060, and 2080, and to explore the spatial clustering effects and risk driving factors. The results show that over 80% of cities are projected to experience considerable increases in flooding volume and economic losses compared to 2020, with higher growth rates in North, Northeast and Southern China. Factors such as urban land compactness and gross domestic product have important impacts on flooding risks. This study offers scientific references for formulating targeted urban flooding management policies and planning.

In the context of global climate change and accelerating urbanization, the increasing frequency and severity of extreme rainfall and flooding events pose significant threats to both natural and human systems. This study aims to develop a Flood Resilience Index (FRI) to quantify and assess environmental resilience to flooding events.

The study adopts the ISO 14090 framework for climate change risk assessment, establishing hazard, exposure, and sensitivity as the three key internal resilience indicators. We employ the 3Di hydrodynamic model to simulate flooding events and derive hazard indicators based on physical flood characteristics in the study area. Additionally, we incorporate habitat connectivity, calculated using Conefor Sensinode 2.2, as a critical factor within the exposure indicator to construct a comprehensive environmental Flood Resilience Index.

Our proposed FRI framework introduces novel perspectives on environmental resilience assessment and provides decision-makers with practical strategies for enhancing adaptive management in flood-prone regions.

The urbanization process significantly alters the hydrological flow patterns of natural watersheds, resulting not only in an increase in liquid flows but also in the amount of eroded sediments transported by surface runoff. In the context of urban flood modeling, the effects of sediment incorporation into the drainage system are generally limited to an increase in the roughness coefficient or the application of a reduction factor in the system's hydraulic capacity. This modeling approach decreases the reliability of extrapolating the results obtained for different situations used in model calibration, as it does not simulate the system clogging process itself but rather the effects it causes. Thus, the objective of this paper is to present the incorporation of a sediment transport and deposition module into MODCEL, a quasi-2D hydrodynamic model, to simulate the transport and deposition of sediments inside urban drainage pipes in real-time simulation, based on the moving bed approach. It was observed that the incorporation of the moving bed did not introduce significant complexities into the calculations nor compromise the model's performance, yielding more reliable results, especially for predicting simulation scenarios.