A Hybrid Deep-Learning Model for El Ni~no Southern Oscillation in the Low-Data Regime

2024-12-04

AI Summary

Addressing the challenges of scarce centennial-scale ENSO observational data, strong nonlinearity, and asymmetric warm/cold phase evolution, this paper proposes a hybrid LIM–deep learning forecasting framework. It pioneers the coupling of a Linear Inverse Model (LIM) with a non-Markovian deep neural network, leveraging LIM’s data efficiency and physical consistency while employing the deep network to correct residual nonlinearities and capture long-range temporal dependencies. Evaluated under data-limited conditions, the hybrid model significantly outperforms both standalone LIM and end-to-end deep learning baselines. It achieves an 18% improvement in forecast skill beyond 9 months, particularly enhancing phase progression and amplitude prediction over the western tropical Pacific. Crucially, it retains LIM’s robustness in identifying strong ENSO events while substantially improving accuracy in both magnitude and timing—establishing a new paradigm for interpretable, high-precision climate prediction under small-sample regimes.

Technology Category

Machine Learning: Large Multimodal Models (LMMs)Natural Language Processing: Learning & Optimization for NLPCognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMsSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingWeb Mining and Content Analysis: Large pretrained models with web data

Abstract

While deep-learning models have demonstrated skillful El Ni~no Southern Oscillation (ENSO) forecasts up to one year in advance, they are predominantly trained on climate model simulations that provide thousands of years of training data at the expense of introducing climate model biases. Simpler Linear Inverse Models (LIMs) trained on the much shorter observational record also make skillful ENSO predictions but do not capture predictable nonlinear processes. This motivates a hybrid approach, combining the LIMs modest data needs with a deep-learning non-Markovian correction of the LIM. For O(100 yr) datasets, our resulting Hybrid model is more skillful than the LIM while also exceeding the skill of a full deep-learning model. Additionally, while the most predictable ENSO events are still identified in advance by the LIM, they are better predicted by the Hybrid model, especially in the western tropical Pacific for leads beyond about 9 months, by capturing the subsequent asymmetric (warm versus cold phases) evolution of ENSO.

Problem

• Hybrid model improves ENSO prediction with limited data
• Combines LIM efficiency and deep-learning nonlinear correction
• Enhances forecast skill for asymmetric ENSO evolution

Innovation

• Hybrid model combines LIM and deep-learning
• Non-Markovian correction enhances LIM predictions
• Captures asymmetric ENSO evolution better