Report on Methods for Quantifying the Dynamics and Feedbacks of Multi-Hazard Risk

Real-world disasters show that there are many dynamics and feedbacks at play that are notcaptured by traditional risk models. MYRIAD-EU’s WP4 aims to improve modelling multi-(hazard)risk dynamics and feedbacks. In this deliverable report, we demonstrate qualitative andquantitative methodologies that we have used to quantify dynamics and feedbacks of multi-hazardrisk and that provide insights into changes over time.Our WP4 research explores various approaches, including machine learning techniques to capture multi-hazard dynamic footprints and susceptibility maps and statistics to compare single- and multi-hazard impacts. We also developed dynamic vulnerability functions using Generalized Additive Models (GAM) and supervised machine learning to assess impacts on water quality and coastalareas. Additionally, qualitative methodologies such as interviews and forensic studies provideinsights into changes in vulnerability and exposure. We investigate vulnerability indicators torepresent multi-hazard dynamics, revealing spatial patterns of social vulnerability and developingprofiles for different phases of the disaster risk management cycle. Furthermore, we introduce aquantitative methodology using nighttime light satellite data to compare and characterise recoveryfrom single and multi-hazard events.Future work by WP4 will refine these methods through local-to-global studies and develop an onlinedatabase to provide empirical evidence of multi-hazard risk dynamics. Our findings aim to supportdecision-makers and involve communities, enhancing resilience through pilot case studies andsectoral collaborations.