Understanding resilience to High-Impact Low-Probability events: a tiered stress testing methodology implemented in the municipality of Venice 

High-Impact Low-Probability (HILP) events pose a growing challenge for contemporary risk assessment and management. These events are characterized by severe consequences, systemic disruptions and limited historical precedent, which constrains the applicability of conventional probabilistic risk assessment methods. As a result, HILP events are often underestimated or excluded from standard decision-making processes. At the same time, their frequency is increasing due to rising interconnections among social, ecological, and infrastructural systems. which amplify the potential for cascading and non-linear effects, allowing disruptions originating in one sector or location to propagate rapidly across multiple domains.In response to these challenges, the AGILE project develops a comprehensive methodology to understand, assess, manage and communicate HILP events through a systemic, risk-agnostic and resilience-oriented framework. Rather than focusing on individual hazards or isolated assets, AGILE conceptualizes risk as an emergent property of interacting systems and emphasizes the capacity of territories to absorb shocks, maintain critical functions and adapt to future conditions. The methodology is structured into three tiers of increasing analytical depth, enabling progressive refinement as data availability, modelling capacity and stakeholder engagement evolve. The framework is applied to the Metropolitan City of Venice, a highly relevant testbed due to its exposure to multi-hazard risks and relevant interdependencies among environmental systems, cultural heritage, tourism, infrastructure, and governance. The first Tier consisted in a workshop-based approach involving experts from different sectors of society. Participants engaged in a serious game designed to qualitatively identify shared vulnerabilities and critical points. During the game, a catastrophic scenario was created based on hazards randomly drawn from a specially designed card deck. Participants analyzed the scenario, attempting to anticipate potential cascading dynamics and identify common points of failure. These exercises encourage lateral and systems-oriented thinking, with a strong focus on cascading effects across interconnected functions and sectors.Building on the outcomes of the first Tier, the second Tier focused on physical and decision-making interdependencies among critical infrastructures and on cascading effects. Through a dedicated workshop, stakeholders translated qualitative insights into interdependency matrices that capture the strength and direction of interactions among infrastructures. These matrices were then used to construct a simplified network of structural and functional dependencies, enabling the identification of systemic vulnerabilities, bottlenecks, and critical nodes, and supporting the analysis of potential cascading failures within the urban system. The third Tier develops a spatial, dynamic, quantitative model of selected critical functions within the municipality, showing how the risk can propagate through the system and evaluates possible resilience improvements by adopting methods such as Network Sciences, Agent Based Modelling and Machine Learning.The system is represented as a network of networks, in which critical functions, such as mobility, energy, and water management, are modelled as interconnected units. This representation enables the analysis of both localized disruptions and their propagation across the wider urban and regional system.Overall, the case study demonstrates how the AGILE framework supports a systemic understanding of cascading effects and resilience pathways under HILP conditions, providing a robust foundation for resilience-oriented planning and decision-making.