Extreme events are difficult to understand because observations are sparse in both space and time, especially for the less frequent but most impactful and severe events. Inference on environmental extremes is becoming increasingly challenging as their behavior can change over time, they can occur across large spatial regions, and they can interact with one another. Extreme Value Theory (EVT) provides a strong theoretical framework for studying these events, but challenges remain in linking methodological advances with environmental applications, prediction, and risk assessment.
We welcome contributions spanning theory, methodology, and applications, including but not limited to:
Advancing EVT methods
• Extreme quantile regression and estimation, conformal prediction, and non-stationary tail models
• EVT-constrained machine learning, uncertainty quantification, extrapolation, and inference under limited extreme event observations
• Methods to evaluate the appropriateness of EVT estimates, impacts of violations of EVT assumptions, and alternative extreme value analysis approaches
Applying EVT to environmental extremes
• Applications to hydrology, climate, weather, coastal hazards, earthquakes, landslides, wildfires, and infrastructure related risk
• Compound, connected, and cascading extremes, including drought-flood sequences, heatwave clusters, extreme precipitation, storm surges, and other interacting hazards
• Novel and under explored applications of EVT to environmental processes and hazards, including snowmelt, air turbulence, environmental epidemiology, and other emerging areas
EVT for prediction, risk assessment, and decision support
• Synthetic extreme event generation and scenario design for stress testing
• Tail focused calibration, validation, and verification, including extremal scoring rules, return level skill, and reliability in the tails
• Quantification and communication of uncertainty relevant to hazards, exposure, impacts, and risk
• Approaches that translate tail behavior and return level information into decision relevant metrics and services
We encourage contributions that bridge methodological EVT developments with real world environmental applications. Contributions may include new theoretical or methodological developments, open datasets and tools, model evaluation frameworks, and real world case studies.