Data supporting the publication "Unrealised extreme storylines of the 2023 North Atlantic marine heatwave revealed through ensemble boosting"

Unrealised extreme storylines of the 2023 North Atlantic marine heatwave revealed through ensemble boosting This Zenodo record accompanies the paper “Unrealised extreme storylines of the 2023 North Atlantic marine heatwave revealed through ensemble boosting” by Catherine H. Gregory, Urs Hofmann Elizondo, Thibault Guinaldo, and Thomas L. Frölicher. Overview This record contains the GitHub repository and associated simulation output used for the study. The project investigates the 2023 North Atlantic marine heatwave and explores physically plausible but unrealised extreme storylines using ensemble boosting experiments. Repository contents The GitHub repository contains analysis scripts, data products, and ensemble-boosting tools used in the study. It includes Jupyter notebooks for analysing and visualising marine heatwave evolution, sea-surface temperature anomalies, subsurface heatwave structure, and heat-budget diagnostics. It also contains filtered marine heatwave event datasets for boosted and double-boosted North Atlantic experiments. Simulation data The simulation output is provided as a compressed archive named Ensemble_boosting_simulations.zip. The model output in this archive has been reduced to the variables and broader study region relevant for the analyses presented in the paper. Ensemble boosting module A dedicated ensemble_boosting module provides tools to generate small, spatially distributed atmospheric temperature perturbations for climate-model ensemble branching. These perturbations are constructed with zero area-weighted global mean, enabling reproducible ensemble-boosting experiments while preserving global energy consistency. The module includes command-line utilities for creating perturbation masks, recording random seeds, and integrating perturbations into atmospheric restart files. Purpose Together, the repository and reduced simulation output support the analysis workflow used to identify, characterise, and visualise unrealised but physically plausible extreme storylines of the 2023 North Atlantic marine heatwave under ensemble-boosted climate-model simulations. Citation Gregory, C. H., Hofmann Elizondo, U., Guinaldo, T., & Frölicher, T. L. Unrealised extreme storylines of the 2023 North Atlantic marine heatwave revealed through ensemble boosting. Corresponding author: Catherine Gregory, [e-mail]