A Generative Likelihood Framework for High-Resolution Climate Model Evaluation

ICLR 2026 Conference Submission19635 Authors

Keywords: High-resolution climate models, climate model evaluation, normalising flows, deep generative models

TL;DR: We introduce a novel approach to quantitatively evaluate high-resolution climate simulations against satellite observations, using the distance between generative likelihood distributions to quantify similarity.

Abstract: Next-generation high-resolution (km-scale) climate models promise unprecedented accuracy in climate projections, but realising their potential requires robust methods to quantify how well simulations align with real-world observations. Average-based metrics conventionally used for climate model evaluation ignore the physics encoded in the finescale structures of km-scale simulations. To overcome this limitation, we propose a novel, statistically principled evaluation methodology based on the likelihood function of a generative image model. Our method provides a continuous similarity metric derived from the likelihood distribution of observation and simulation snapshots, which can redefine the evaluation, intercomparison, and parameter tuning of high-resolution climate models. We demonstrate the applicability and interpretability of this method by evaluating convective clouds simulated by two state-of-the-art global km-scale models, using their outgoing infrared radiation fields. This work establishes a scalable pathway toward observation-based evaluation of next-generation climate simulations.

Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)

Submission Number: 19635