Explore the Ideology of Deep Learning in ENSO Forecasts
This work addresses the opacity of deep learning models in climate science, which hampers trust and deployment, though it is incremental in improving interpretability for a specific domain.
The paper tackled the challenge of interpreting deep learning models for El Niño-Southern Oscillation (ENSO) forecasts by introducing a mathematically grounded interpretability framework, which revealed that ENSO predictability primarily stems from the tropical Pacific with contributions from other oceans, and identified that the Spring Predictability Barrier persists due to suboptimal variable selection.
The El Ni{~n}o-Southern Oscillation (ENSO) exerts profound influence on global climate variability, yet its prediction remains a grand challenge. Recent advances in deep learning have significantly improved forecasting skill, but the opacity of these models hampers scientific trust and operational deployment. Here, we introduce a mathematically grounded interpretability framework based on bounded variation function. By rescuing the "dead" neurons from the saturation zone of the activation function, we enhance the model's expressive capacity. Our analysis reveals that ENSO predictability emerges dominantly from the tropical Pacific, with contributions from the Indian and Atlantic Oceans, consistent with physical understanding. Controlled experiments affirm the robustness of our method and its alignment with established predictors. Notably, we probe the persistent Spring Predictability Barrier (SPB), finding that despite expanded sensitivity during spring, predictive performance declines-likely due to suboptimal variable selection. These results suggest that incorporating additional ocean-atmosphere variables may help transcend SPB limitations and advance long-range ENSO prediction.