An optimal path threshold method for rigorously identifying extreme climate events

Abstract

The frequent occurrence of extreme hydroclimate events has a significant impact on ecosystems and socio-economic systems. However, intercomparisons of extreme events are often hampered by inconsistent definitions and standardizations for different types of events. Here we introduce the Optimal Path Threshold (OPT) method, a dynamic programming-based approach that adaptively determines optimal thresholds for defining extreme events at specific severity levels. Two case studies demonstrate its applicability. The first case study presents a long-term comparison of air heatwaves and lake heatwaves in Lake Mead. The algorithm adaptively selects thresholds to identify a comparable number of events at various severity levels. The second case study examines compound hot, dry, and windy events in California, exploring threshold selection for multivariable extremes and investigating their potential relationship with wildfires. The OPT method proves effective for adaptably and objectively analyzing complex and multifactorial extreme events.

Keywords

• climate change
• extreme events
• heatwave