Multi-Risk Dynamics

In our increasingly unpredictable world, understanding how different hazards interact and affect the risks we face is crucial for better disaster preparedness and response. Our work aims to build an evidence base to grasp the complex dynamics between various risk factors, known as risk drivers, and how they are influenced by multi-hazard events. These events could be anything from consecutive storms to co-occurring droughts and heatwaves, amplifying the overall risk. To achieve this, the team employs a range of innovative techniques such as machine learning and disaster forensics, alongside traditional methods, to analyze past incidents and detect changes in vulnerability and exposure over time and space. They also conduct interviews to gather empirical evidence from pilot regions, enriching their understanding of dynamic feedbacks. The insights gained from this research are then used to develop methods and tools to explicitly represent these dynamic interactions in risk models. The ultimate goal is to create a comprehensive online database that provides real-world examples and data on how hazards interact with risk factors, aiding policymakers, planners, and communities in making informed decisions and building resilience against future disasters. 

Introduction

The work in the multi-risk dynamics theme focuses on investigating the complex interactions and feedback loops between various risk drivers in the context of multi-hazard events. The primary objective is to develop an evidence-based understanding of how different hazards, such as floods, storms, and earthquakes, influence the dynamics of risk factors like vulnerability and exposure over time and space.

To achieve this, the project employs a combination of novel and traditional methods. Novel approaches include machine learning techniques, disaster forensics analysis, and the integration of diverse data sources such as disaster loss databases, national statistics, social media, and night-time light data. These methods are utilized to detect changes in reported losses and damages attributed to spatial and temporal variations in exposure and vulnerability resulting from multi-hazard events. Additionally, we developed an interview methodology to extract empirical evidence from sector-specific pilots, enhancing the understanding of dynamic feedbacks within and across specific sectors.

The insights obtained from these analyses are then used to develop functions and methods for explicitly representing dynamic feedbacks between risk drivers in risk modelling approaches. These methods aim to improve the accuracy of forward-looking multi-risk scenarios, thereby enhancing disaster preparedness and response efforts. Furthermore, the project aims to create an online database that provides empirical evidence and data on short-to-long term changes in exposure and vulnerability due to interactions with hazards. This database will be integrated into the MYRIAD-EU dashboard, providing stakeholders with valuable insights for decision-making and risk management strategies.

Application of Methods and Tools within Framework

Finding a system definition

To support finding a system definition, we developed an interview methodology. On the one hand, we designed the interviews to obtain information on multiple aspects:

  • Find out about the interviewees' awareness of various combinations of hazards impacting their region, including whether these combinations are becoming more significant and are explicitly considered in their work which can aid in the identification of relevant single and multi-hazards scenarios.
  • Get insights in the key determinants of vulnerability to different hazards, exploring how these characteristics vary across sectors and sometimes deviate from initial perceptions, ultimately aiding in identifying exposed elements and broader vulnerabilities within the system for effective risk management and resilience planning.
  • Understand how actions taken to mitigate one hazard might positively or negatively affect other hazards in the region, illustrating scenarios where risk reduction measures aimed at one hazard inadvertently impact another, thus informing more comprehensive risk management strategies by considering potential unintended consequences. This can provide useful examples of management challenges and solutions

We are also developing an approach for creating multi-hazard susceptibility maps which provide spatial information on the various natural hazards which potentially occur simultaneously or sequentially for aregion of interest helping decision makers to select relevant hazards and hazard combinations for their assessment.

In addition, we developed VulneraCity, a comprehensive database, which offers a wealth of information on urban vulnerability drivers for various hazards like coastal flooding and earthquakes, aiding in the identification of exposed elements within the system, such as buildings or vulnerable populations, and providing insights into broader vulnerability aspects such as economic and social disparities, thereby facilitating more informed risk assessment and mitigation strategies.

Finally, we are building a multi-hazard events database which includes information on implemented risk management options and their performance in 100+ multi-hazard cases that can serve as examples.

Characterisation of direct risk

To aid with the identification of direct impacts, the multi-hazards events database of 100+ multi-hazard events offers a catalogue of past disasters to including information on direct impacts that can serve as reference cases.

To assess direct risk metrics we have developed three approaches:

  1. We are developing heat related human impact functions to pinpoint impact-relevant durations and hazard-impact thresholds by considering factors such as temperature for heatwaves and soil moisture for droughts
  2. We are developing a framework for artificial intelligence for climate change multi-risk assessment to identify and analyse multi-hazard events, helping to understand their impacts and vulnerabilities, particularly demonstrated in the assessment of extreme climate events in the Veneto Region of North-East Italy, enabling more effective disaster planning and decision-making.
  3. We are developing a nighttime light satellite data approach to estimate the duration of post-event recovery for single-and multi-hazard events

To help identify and understand dynamics of exposure, vulnerability and impacts, we have developed and are developing three different tools:

  1. We have developed the interview methodology to discover changes in exposure and vulnerability characteristics, which aims to comprehend how socio-economic shifts, prolonged disasters, and multi-hazard scenarios have altered vulnerability or exposure conditions in the region, providing examples such as urban development increasing flood exposure or drought management strategies inadvertently exacerbating flood vulnerabilities, thus informing risk assessment by understanding how hazards impact various elements over time and space.
  2. We are developing the multi-hazards events database which describes dynamics of exposure, vulnerability and impacts 100+ multi-hazard events and can be helpful reference for other cases.
  3. We are preparing insights into how social vulnerability indicators in the disaster risk reduction cycle are pivotal throughout the disaster risk reduction cycle, aiding in preparedness by identifying at-risk populations, guiding resource allocation in response, informing recovery fund allocation, and directing targeted interventions in mitigation efforts. We explain the relevance and provide examples of how this can reveal dynamics of vulnerability.

Defining risk management options

To support selecting risk management options we have developed and are developing two approaches:

  1. We designed the interview methodology to investigate how actions taken to mitigate one hazard might positively or negatively affect other hazards in the region, illustrating scenarios where risk reduction measures aimed at one hazard inadvertently impact another, thus informing more comprehensive risk management strategies by considering potential unintended consequences. These insights can help evaluate risk management options.
  2. The multi-hazards events database will include information on implemented risk management options and their performance in 100+ multi-hazard cases which can be valuable reference for other cases.