LUCIE: A Lightweight Uncoupled ClImate Emulator with long-term stability and physical consistency for O(1000)-member ensembles

2024-05-25

arXiv.org

AI Summary

This work addresses the lack of lightweight, low-data-dependency, and long-term stable fully data-driven climate simulators. We propose LUCIE, a novel simulator trained on only two years of 6-hourly ERA5 data. LUCIE employs a low-resolution neural network architecture, autoregressive modeling, spectral-domain regularization, and—crucially—a novel hard-constrained first-order integrator to suppress error accumulation. A low-data-dependency optimization algorithm enables efficient training in just 2.4 hours on a single A100 GPU. LUCIE supports large-ensemble (thousand-member), century-scale autoregressive simulations while accurately reproducing ERA5’s long-term climatology, interannual variability, and extreme-event return periods across temperature, wind, humidity, and precipitation fields. It ensures both physical consistency—via spectral constraints and integrator design—and numerical stability over extended integration. Thus, LUCIE establishes an efficient, robust, data-driven paradigm for large-scale ensemble climate simulation.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Search and Optimization: Sampling/Simulation-based Search
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior

Application Category

• Search and Retrieval-Augmented AI: Large language models for search
• Web Mining and Content Analysis: Web data generation and simulation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

We present a lightweight, easy-to-train, low-resolution, fully data-driven climate emulator, LUCIE, that can be trained on as low as $2$ years of $6$-hourly ERA5 data. Unlike most state-of-the-art AI weather models, LUCIE remains stable and physically consistent for $100$ years of autoregressive simulation with $100$ ensemble members. Long-term mean climatology from LUCIE's simulation of temperature, wind, precipitation, and humidity matches that of ERA5 data, along with the variability. We further demonstrate how well extreme weather events and their return periods can be estimated from a large ensemble of long-term simulations. We further discuss an improved training strategy with a hard-constrained first-order integrator to suppress autoregressive error growth, a novel spectral regularization strategy to better capture fine-scale dynamics, and finally an optimization algorithm that enables data-limited (as low as $2$ years of $6$-hourly data) training of the emulator without losing stability and physical consistency. Finally, we provide a scaling experiment to compare the long-term bias of LUCIE with respect to the number of training samples. Importantly, LUCIE is an easy to use model that can be trained in just $2.4$h on a single A-100 GPU, allowing for multiple experiments that can explore important scientific questions that could be answered with large ensembles of long-term simulations, e.g., the impact of different variables on the simulation, dynamic response to external forcing, and estimation of extreme weather events, amongst others.

Problem

• Emulate climate with long-term stability and physical consistency
• Train lightweight model with minimal data (2 years)
• Estimate extreme weather events using large ensembles

Innovation

• Hard-constrained first-order integrator suppresses error growth
• Novel spectral regularization captures fine-scale dynamics
• Optimization enables data-limited training without losing stability