Innovative Universal Severity Classification System for Natural Disasters

Abstract

Assessing the true impact of natural disasters is complex, varying by type, location, and time period. Natural disasters, ranging from minor fires to cataclysmic super-volcanic eruptions, universally affect living beings and their habitats. No single factor captures all aspects of impact, as different factors convey varying perspectives on severity, and not all factors are equally significant. Disasters are frequently categorized based on criteria like fatalities or financial damage, but existing classifications often lack a scientific foundation to capture the overall impact, leading to inconsistent measurements and management. The absence of a coherent, globally accepted standard for communicating disaster severity complicates effective disaster management, as varied interpretations hinder unified response efforts. Diverse scales, such as the EF-Scale for tornadoes and the VEI scale for volcanic eruptions, are used to measure disaster magnitude. Various stakeholders—emergency responders, humanitarian aid workers, medical personnel, insurance managers, media, and civilians—employ different scales and terminologies, creating coordination challenges. Disconnected measurement systems cause confusion, hinder relief and mitigation efforts, delay resource allocation, and can lead to political and social instability, affecting international security and relations and causing long-term negative impacts for affected regions. To address these issues, this dissertation introduces the Dis-Utility-Based Universal Disaster Severity Classification System (DUDSCS), a novel framework for assessing and comparing the severity of natural disasters. The DUDSCS integrates both quantitative and qualitative criteria, the former employing 'dis-utility functions' to identify key severity factors and their relationships. Factors such as fatalities, injuries, homelessness, affected populations, and economic damage are evaluated to provide a comprehensive measure of event severity. The development of the DUDSCS involved a six-step methodology, starting with the adoption of 'random dis-utility theory' derived from choice modeling in random utility theory. A global survey captured perceptions, resulting in over 6,000 responses used to calibrate and validate the dis-utility model. Additionally, an in-depth analysis of over 13,000 historical disaster records and their probability distribution helped establish the system's quantitative criteria. By offering a consistent, multidimensional assessment tool, the DUDSCS enhances communication and coordination among stakeholders. By addressing inconsistencies in disaster terminologies and providing a common communication platform, the DUDSCS promotes better-prepared, more resilient communities. The qualitative classification system, while originally in English, can be adapted to other languages using clear definitions for each level based on impact degree, circumstances, fatalities, injuries, and damage. A color-coding system assists disaster recovery workers who may be illiterate or unfamiliar with the local language, enabling effective communication and response in foreign regions. This ensures that all individuals, regardless of language or literacy barriers, can understand and respond to disasters efficiently. This scale captures varying levels of severity, making it a valuable tool for diverse stakeholders, including policymakers, disaster managers, insurance estimators and researchers. The DUDSCS framework provides a new 0 to 10 scale for event severity, where zero indicates minor impact and ten signifies partial or full extinction. The research includes a case study on the COVID-19 pandemic, demonstrating the system's applicability and addressing key gaps in disaster literature. The system aligns with global frameworks such as the Sendai Framework for Disaster Risk Reduction 2015-2030, contributing to targets related to reducing disaster impacts on lives, economies, and infrastructure. Despite challenges such as data biases and predictive limitations, this comprehensive, statistically sound DUDSCS marks a significant advancement in disaster impact assessment and management, fostering informed decision-making and effective resource allocation for disaster preparedness and response.