ClimateLLM: Efficient Weather Forecasting via Frequency-Aware Large Language Models

2025-02-16

AI Summary

Existing meteorological forecasting models face three key challenges: (1) difficulty in modeling dynamic temporal dependencies and short-term abrupt changes—hindering extreme weather prediction; (2) high computational overhead; and (3) weak adaptability to multi-scale frequency-domain patterns. To address these, we propose Freq-LLM, a frequency-aware meteorological large language model. Freq-LLM introduces a novel Fourier frequency-domain decomposition framework tightly integrated with LLMs; designs a frequency-adaptive Mixture-of-Experts (MoE) mechanism to jointly model global trends and local anomalies; and incorporates cross-spatiotemporal dynamic meteorological prompts to enhance physical consistency. Evaluated on real-world meteorological datasets, Freq-LLM significantly outperforms state-of-the-art methods, achieving substantial gains in forecast accuracy, 37% faster inference speed, and scalable global-scale forecasting capability.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Weather forecasting is crucial for public safety, disaster prevention and mitigation, agricultural production, and energy management, with global relevance. Although deep learning has significantly advanced weather prediction, current methods face critical limitations: (i) they often struggle to capture both dynamic temporal dependencies and short-term abrupt changes, making extreme weather modeling difficult; (ii) they incur high computational costs due to extensive training and resource requirements; (iii) they have limited adaptability to multi-scale frequencies, leading to challenges when separating global trends from local fluctuations. To address these issues, we propose ClimateLLM, a foundation model for weather forecasting. It captures spatiotemporal dependencies via a cross-temporal and cross-spatial collaborative modeling framework that integrates Fourier-based frequency decomposition with Large Language Models (LLMs) to strengthen spatial and temporal modeling. Our framework uses a Mixture-of-Experts (MoE) mechanism that adaptively processes different frequency components, enabling efficient handling of both global signals and localized extreme events. In addition, we introduce a cross-temporal and cross-spatial dynamic prompting mechanism, allowing LLMs to incorporate meteorological patterns across multiple scales effectively. Extensive experiments on real-world datasets show that ClimateLLM outperforms state-of-the-art approaches in accuracy and efficiency, as a scalable solution for global weather forecasting.

Problem

• Improves weather prediction accuracy
• Reduces computational costs significantly
• Adapts to multi-scale frequency variations

Innovation

• Frequency-Aware Large Language Models
• Mixture-of-Experts mechanism
• Cross-temporal dynamic prompting mechanism