Climate change is accelerating the emergence of interconnected natural hazards, increasing not only the frequency and intensity of extreme events but also the occurrence of cascading processes in which one hazard modifies environmental conditions and amplifies the likelihood of subsequent hazards. Prolonged droughts alter vegetation structure and fuel moisture, increasing wildfire susceptibility, while wildfire-induced changes in soil properties, land cover, and watershed response can substantially intensify post-fire flooding, debris flows, and erosion. These complex interactions challenge traditional single-hazard approaches and call for integrated frameworks capable of capturing the dynamic relationships among climate, landscape evolution, and hydrological processes.

This plenary presents recent advances from our research on AI-driven decision support for cascading natural hazards, with a particular focus on the drought–wildfire–flood continuum. The work integrates multi-source Earth observation data, remote sensing, climate information, topographic and hydrological datasets, and environmental variables within advanced machine learning and GeoAI frameworks to improve hazard susceptibility mapping, wildfire prediction, flood forecasting, and spatial risk assessment. Rather than viewing these hazards independently, the proposed approaches explicitly account for their interactions and cumulative impacts across space and time.

A central theme of the presentation is the development of trustworthy AI for environmental decision-making. The talk will demonstrate how Explainable Artificial Intelligence (XAI), uncertainty quantification, Bayesian optimization, ensemble learning, and reinforcement learning can enhance both predictive performance and scientific understanding by identifying the dominant environmental drivers, nonlinear relationships, and spatial dependencies governing hazard evolution. These advances transform AI from a black-box prediction engine into an interpretable and reliable decision-support tool.

Beyond prediction, the presentation highlights the evolution of AI toward operational intelligence. Examples from recent research will illustrate how predictive models can support early warning, resource allocation, mitigation planning, infrastructure resilience, and climate adaptation by delivering transparent, actionable information to emergency managers, planners, and policymakers. The integration of explainability, optimization, and uncertainty analysis enables more informed and defensible decisions under rapidly changing environmental conditions. Moving beyond isolated hazard prediction toward holistic, explainable, and operational AI frameworks offers a new paradigm for managing cascading natural hazards and strengthening societal resilience in an era of accelerating climate change.

Dr. Hossein Bonakdari, Ph.D., P.Eng., is a Professor in the Department of Civil Engineering at the University of Ottawa, Canada, and an internationally recognized expert in Artificial Intelligence for environmental and climate systems. His research lies at the intersection of artificial intelligence, geospatial analytics, hydrology, hydraulics, remote sensing, and environmental data science, with a focus on developing next-generation AI solutions for climate resilience, water resources management, and natural hazard assessment.
Over the past two decades, Dr. Bonakdari has pioneered the development of physics-informed AI frameworks for complex environmental systems. His research has advanced machine learning, deep learning, GeoAI, and hybrid intelligence approaches for flood forecasting, drought assessment, wildfire susceptibility modeling, hydro-meteorological prediction, climate-risk analysis, and AI-driven decision-support systems. By bridging fundamental AI research with operational applications, his work has enabled more transparent, reliable, and actionable tools for environmental management, disaster resilience, and sustainable infrastructure planning.

Dr. Bonakdari has authored more than 350 peer-reviewed publications, including eight books, over 30 book chapters, and has delivered more than 150 keynote, invited, and conference presentations at international scientific meetings. His publications have received over 11,000 citations, and he has an H-index of 62, reflecting the broad impact of his research across civil engineering, environmental sciences, hydrology, and artificial intelligence. He has been recognized among the World's Top 2% Scientists since 2019, highlighting his sustained scientific excellence and international leadership in AI-driven environmental research.