Accurately modeling extreme weather events is a major goal in climate science and weather risk management. High-resolution computational models provide precise data but are costly for capturing rare events. This study focuses on localized, short-lived events like heavy precipitation. The researchers, including Paul O’Gorman, developed a more efficient method for sampling these events by combining adaptive multilevel splitting (AMS) and “ensemble boosting.” The new approach, tested on the Lorenz-96 model, improved the sampling of extreme events by a factor of 10. This method helps efficiently simulate sudden, transient weather extremes, advancing efforts to model these events in atmospheric models.