Capturing Unseen Spatial Heat Extremes Through Dependence‐Aware Generative Modeling
Open Access
- 19 September 2026
- journal article
- research article
- Published by American Geophysical Union (AGU) in Earth's Future
- Vol. 14 (9)
Abstract
Observed records of climate extremes provide an incomplete view of plausible hazards, missing “unseen” events beyond historical experience. Ignoring spatial dependence further underestimates hazards striking multiple locations simultaneously. We introduce DeepX‐GAN (Dependence‐Enhanced Embedding for Physical eXtremes—Generative Adversarial Network), a deep generative model that explicitly captures the spatial structure of rare extremes. Its zero‐shot generalizability enables the simulation of statistically plausible extremes beyond the observed record, evaluated against long climate model large‐ensemble simulations. We define two unseen types: direct‐hit extremes that affect the target, and near‐miss extremes that narrowly miss. These unrealized events reveal hidden risks and can either prompt proactive adaptation or reinforce a false sense of resilience. Applying DeepX‐GAN to the Middle East and North Africa shows that the probability of unseen heat extremes is disproportionately distributed toward countries with high vulnerability and low socioeconomic readiness. Using a representative climate simulation, we demonstrate how future warming could expand and shift these hazards, creating persistent hotspots in Northwest Africa and the Arabian Peninsula and new hotspots in Central Africa, necessitating spatially adaptive resilience planning.Keywords
All Related Versions
This publication has 45 references indexed in Scilit:
- Downscaling with AI reveals the large role of internal variability in fine-scale projections of climate extremesClimate Dynamics, 2026
- Climate model large ensembles as test beds for applied compound event researchiScience, 2024
- Generative Adversarial Networks2023
- Deep Generative Modeling2022
- Changes in climate extremes in observations and climate model simulations. From the past to the future2020
- Multivariate Copula Analysis Toolbox (MvCAT): Describing dependence and underlying uncertainty using a Bayesian frameworkWater Resources Research, 2017
- Reconceptualising adaptation to climate change as part of pathways of change and responseGlobal Environmental Change, 2014
- Methods of Tail Dependence Estimation2012
- Extremes in Nature2007
- An Introduction to Statistical Modeling of Extreme Values2001