Conference Agenda

Flooding is a common disaster worldwide, having enormous detrimental impacts on society. The frequency and intensity of floods are rising due to climate change and the consequential damage is dramatically increasing as a result of elevated exposure. With the increasing frequency and intensity of flooding, it has become imperative to mitigate flood risks effectively. A great amount of research is focused on risk assessment and mitigation strategies. The effects can be significantly reduced through the implementation of risk mitigation strategies including providing decision-makers with relevant, accurate risk assessment information that can aid in establishing effective emergency management protocols and communicating with the public to save lives and property. Flood risk assessment contributes significantly to managing floods effectively. These assessments enable us to identify possible threats at both the global and local levels, providing vital information for mapping high-risk areas. Machine Learning (ML) models can simulate flood risk by identifying critical non-linear relationships between flood damage locations and flood risk factors (FRFs). Tampa Bay, Florida, is selected as a test site. The study's goal is to simulate flood risk and identify dominant FRFs using five ML models: Decision Tree (DT), Support Vector Machine (SVM), Adaptive Boosting (AdaBoost), Extreme Gradient Boosting (XGBoost), and Random Forest (RF). Historical flood damage data is the target variable, with 17 FRFs as predictors. RF and XGBoost outperform other models in accuracy tests. RF classifies 2.23% of Tampa Bay as very high risk and 2.55% as high risk, while XGBoost classifies 3.77% as very high risk and 1.09% as high risk. Key FRFs include low elevation, distance from water bodies, extreme precipitation, and population density. This study demonstrates an effective approach to assessing flood risk by considering a wide spectrum of FRFs using advanced supervised ML algorithms. Through this approach, we were able to identify the zones under very high and high flood risks in Tampa Bay, along with the identification of responsible FRFs, in addition to potential exposed populations. Such risk identification procedures will enable us to take precautions and to better prepare for floods at both household and institutional levels. The risk identification is believed to ultimately advance flood resilience. All of the findings contribute to flood risk management both scientifically and practically. The findings can be meaningful in the vast scientific field for flood management studies around the world and the approach will directly assist the policymakers during their decision-making regarding flood risk management.

The increasing complexities of climate change and natural hazard risks pose significant challenges for communities globally, necessitating a deeper understanding of how these risks are perceived at both societal and community levels. The negative impacts of natural hazards and climate change are often exacerbated when these risks are not adequately integrated into land use planning. Currently, New Zealand lacks a comprehensive national framework to establish tolerable risk thresholds related to life, property, and the economy. While several local councils have attempted to develop their own risk-based frameworks, these efforts are constrained by several factors: insufficient community awareness about acceptable risk levels, limited understanding and motivation among decision-makers to adopt risk-based approaches, and inherent uncertainties associated with natural hazards and climate change.

This research explores the complex relationship between individuals' perceptions of climate change and natural hazard risks, examining how these perceptions are shaped by diverse contexts and backgrounds in the societal, geographical and cultural environments. It specifically investigates the factors that influence community perceptions of these risks, focusing on the level of awareness within communities about the significance of these risks and how this awareness affects their actions in managing and mitigating them.

The literature on risk perception in New Zealand, situated within a global context, suggests that risk perception is influenced by various factors, including environmental and cultural characteristics, with significant regional variations (Fan et al., 2022). Research indicates that risk perception is shaped by social representations of the associated risks and that personal values significantly impact attitudes towards climate change (Han et al., 2022). In New Zealand, social vulnerability indicators have been developed to identify populations at risk of flooding—a common natural hazard in the region. These indicators assess community resilience by considering factors such as exposure, health and disability status, social connectedness, and housing (Mason et al., 2021). Moreover, public perceptions of risk often diverge from expert assessments, which are based on empirical data and specialised knowledge (Xiao et al., 2023). Factors influencing risk perception include gender, age, race, education, region of residence, cultural background, social status, income, knowledge of risks, disaster experience, and communication strategies (Xiao et al., 2023).

The Henderson-Massey Local Board area in the Auckland region in New Zealand was selected as a case study, with data collected through both direct household surveys and online questionnaires to assess community ‘understanding’ and ‘awareness’ of their exposure to natural hazards and climate change risks. Preliminary findings highlight a complex interaction of cognitive and social factors that shape these perceptions, subsequently influencing community decision-making and behaviour concerning risk mitigation.

This study advances the field of risk management by integrating community perspectives and providing new insights and developments relevant to both local and international contexts in mainstream climate change adaptation. The findings offer valuable implications for future policy and planning initiatives aimed at enhancing urban resilience and adaptive capacity to natural hazards and climate change exacerbated impacts.

In the complexity of international poly-crisis phenomena, the emergency response to earthquakes involves a multifaceted process, from immediate relief to long-term reconstruction. Geoinformation, cartographic, and aerial data are essential in Disaster Risk Reduction, Disaster Management, and Building Damage Assessment. However, documenting early recovery architectures remains challenging during the in-between period and the reconstruction transition, despite advancements in knowledge platforms, data management, and rapid multiscalar information sharing encouraged by international scientific programs such as the Integrated Research on Disaster Risk and the Sendai Framework for Disaster Risk Reduction 2015-2030.
Geomatic and photogrammetric techniques are central to documenting the emergency process and projects across territorial, urban, and building scales, defining a common semantic and geometric codification for mapping temporary phenomena.
This research focuses on the long-term spatial impacts of all temporary solutions, which have significantly and permanently altered the post-seismic Italian landscapes, particularly in the regions of Central Italy that have been affected by a succession of catastrophic events in the last 25 years.
The research has revealed the lack of overall cartographic documentation of these interventions, which is responsible for new urban configurations, despite the collective planning, economic and regulatory effort made, and the digital technological potential available today. Existing documentation is incomplete, fragmented, and outdated, necessitating a systematic, evidence-based approach to bridge this gap.
The methodology combines diverse data sources – satellite imagery, UAV-based photogrammetry, and temporal and economic processual data – into a unified multiscale geospatial database organized within the GIS platform. This database serves as the conceptual model for the Operative Atlas, facilitating the visualization and analysis of emergency responses and urban infrastructure.
One key aspect of the research is the implementation of attribute specification in existing official cartographic data, developing a standardized terminology and classification system for mapping temporary interventions, and making them recognizable within the national and international cartography during and after emergencies. By developing a conceptual and logical model, the research defines geometric entities in the GIS data structure to document and digitalize these provisional contexts. UAV point clouds are integrated with non-metric data to ensure detailed descriptions through a multiscalar approach. The methodology was validated in the study of Visso (Macerata), a small historical mountain village in Central Italy, where specific temporary interventions were effectively mapped and
documented. In conclusion, the research adopts an interdisciplinary and integrated approach to overcome the existing gaps in the documentation of post-disaster interventions. Combining architectural studies with geomatic tools, the Operative Atlas is valuable for managing heterogeneous geospatial emergency data. Digital visualization techniques can support different emergency phases and various stakeholders involved in post-disaster development strategies, such as local and national emergency management authorities, urban planners, and practitioners, in facilitating the transition from emergency response to permanent settlement. It offers digital visualization techniques to support, in different emergency phases, stakeholders involved in post-disaster development strategies, including local and national emergency management authorities, urban planners, and practitioners in facilitating the transition from emergency response to permanent settlement.

This work presents our current methodological efforts to explore, extract, visualize, and analyze potentially relevant information captured by the citizens' weather reports.

We started this project at the beginning of 2022 when MeteoSwiss shared with us their collection of citizens' observations on weather conditions reported by users of the MeteoSwiss mobile application. We followed an iterative and participatory design process that can be divided into three main phases. In each phase, we developed prototypes, gathered experts' requirements, designed and ran user experts' studies, and collected domain experts' feedback to improve our solution's features.

In Phase 1, we focused on the research of novel visual representations for the spatio-temporal analysis of only the citizens' reports. In Phase 2, we focused on facilitating the visual comparison of radar data and citizens' reports, exploring reports' images and extra metadata, and re-engineering the tool to make the software more accessible, modular, and robust.

Phase 3 is an ongoing effort, and its main goal is to provide meteorologists with information extracted from the images to estimate the impact and damage on the ground as reported by citizens.

A prototype of our visual tools is available at https://weatherva.ifi.uzh.ch/.

Multidimensional datasets are often represented by layering them within a single map, which can lead to an exaggeration of information. As an alternative representation, this study investigates the use of a small multiple layout for web applications, in which several maps with a common spatial area, each representing a different topic, are arranged next to each other and can be viewed simultaneously.

The area of spatial planning was used as a use case, as many different topics of a common area, which are largely obtained from the ÖREB-cadastre, have to be identified and analysed in the planning process. Spatial planning experts took part in the qualitative usability study, carrying out a fictitious task that is typical for this specialist area. The aim of the study was to analyse how the experts rated the small multiple layout in terms of usability and efficiency. The usability test was analysed using a qualitative content analysis.

The experts were able to quickly gain an overview of the topics and make an efficient cross-comparison, particularly when weighing up interests. They were able to complete their tasks efficiently and were predominantly positive about the layout during and after completing the tasks. Limitations arise if the area under consideration extends beyond the boundaries of the small multiple map window. The study concludes that small multiples layouts are a useful and efficient tool for the simultaneous visual analysis of multiple spatially coincident themes, with potential applications extending beyond spatial planning to other fields that use multidimensional spatial data.

Climate change is an ongoing environmental threat, and its mitigation poses significant societal challenges. Policymakers are tasked to make time-critical decisions rapidly, including hazard forecasting, preparedness, warning, and response to mitigate potentially harmful consequences of climate change. This often involves map-based decision-making with visualized climate data, which are confronted with various sources of uncertainty. Uncertainty is an inherently tricky concept; thus, it is challenging to communicate clearly with decision-makers and the public. Especially in the emotionally charged context of climate change, with segments of the population already showing scepticism towards climate change, the communication of uncertain information in map displays still needs deeper investigation.
Applying a mixed factorial (3x2x2) map-based, online experiment, we aimed to study the visualization of (un)certainty in static climate change forecast maps and how this might interact with map readers' emotions and attitudes. Inspired by the CH2018 Scenarios for the year 2060 (NCCS, 2018), we designed change prediction map stimuli with different climate variables in three versions: without any uncertainty information, with uncertainty visualized as black dots (Figure 1) or lines, using empirically validated depiction guidelines (Retchless & Brewer, 2016). Based on prior research, we chose the term 'certainty' instead of 'uncertainty' in our stimuli (Johannsen et al., 2018). Each map was accompanied by context information on its right-hand side. It either included a graphic character with a fitting narrative (between factor: emotion), intended to elicit an emotional response (Fig. 1a), or it just contained information equivalent to facts (Fig. 1b).

Leveraging the crowd-sourcing platform Prolific.com, we recruited 109 people (f: 53, m = 54, nb = 2; average age = 34 yrs.) to participate in this online study. We balanced the number of climate change sceptical and believing participants (between factors: attitude) using screeners in Prolific. All participants were exposed to 18 maps in random order, without any time pressure. They were asked to select the locations predicted to be most/least affected by climate change shown on the map (out of six given options, A-F in Figure 1), to assess the depicted change certainty and severity, and to indicate their amount of trust in the depicted information, on a scale from least (1) to most (7).
Preliminary results (ART ANOVA) indicate that participants reporting a sceptical climate change attitude rate the severity of the depicted change significantly lower compared to believing participants (F = 5.446, p < .05), irrespective of the emotional context. Regardless of their attitudes, participants, on average, trust maps without any certainty information significantly less (F = 167.872, p < .001) compared to when certainty is visually communicated, irrespective of the certainty visualization type.

Our initial findings not only underline the importance of deriving suitable visualization guidelines to communicate uncertainty climate change forecast maps effectively but also suggest adapting the visual communication method to population segments with varying attitudes towards climate change. Future research should deepen our understanding of emotion's role in the visual communication of climate change forecast maps and how different types of climate change scepticism might interact with uncertainty visualization.