LASSE: Learning Active Sampling for Storm Tide Extremes in Non-Stationary Climate Regimes

2024-12-30

AI Summary

Identifying destructive tropical cyclones and simulating associated storm surges under non-stationary climate conditions remains challenging due to high computational cost and poor generalizability of conventional methods. Method: This paper proposes an online active learning framework specifically designed for storm surge extremes. It integrates surrogate modeling, information-entropy-driven adaptive sampling, online incremental training, and high-fidelity hydrodynamic simulation—introducing the first “extreme-oriented” online active sampling paradigm that enables iterative, self-optimizing model refinement. Contribution/Results: Compared to standard Monte Carlo approaches, the framework substantially alleviates computational bottlenecks while maintaining robust generalization under previously unseen climate scenarios. Evaluated on a large-scale downscaled typhoon catalog, it achieves 100% precision and perfect recall for rare destructive storms using less than 20% of the full-simulation cost, establishing a computationally efficient and scalable paradigm for climate risk assessment.

Technology Category

Machine Learning: Active LearningSearch and Optimization: Sampling/Simulation-based SearchReasoning under Uncertainty: Stochastic Optimization

Application Category

Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applicationsSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingGraph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Identifying tropical cyclones that generate destructive storm tides for risk assessment, such as from large downscaled storm catalogs for climate studies, is often intractable because it entails many expensive Monte Carlo hydrodynamic simulations. Here, we show that surrogate models are promising from accuracy, recall, and precision perspectives, and they ``generalize"to novel climate scenarios. We then present an informative online learning approach to rapidly search for extreme storm tide-producing cyclones using only a few hydrodynamic simulations. Starting from a minimal subset of TCs with detailed storm tide hydrodynamic simulations, a surrogate model selects informative data to retrain online and iteratively improves its predictions of damaging TCs. Results on an extensive catalog of downscaled TCs indicate a 100% precision retrieving the rare destructive storms training using less than 20% of the simulations as training. The informative sampling approach is efficient, scalable to large storm catalogs, and generalizable to climate scenarios.

Problem

• Tropical Cyclones
• Storm Surges
• Climate Change Impact

Innovation

• Online Learning Method
• Extreme Storm Surges Prediction
• Efficiency and Accuracy Improvement