Multidata Causal Discovery for Statistical Hurricane Intensity Forecasting

2025-10-02

AI Summary

Statistical intensity forecasting of Atlantic hurricanes suffers from poor generalizability, confounding variable interference, and difficulty in identifying nonlinear causal relationships. Method: We propose a multi-dataset causal discovery–driven feature selection framework that integrates operational SHIPS forecasts with ERA5 reanalysis data—departing from conventional correlation- or goodness-of-fit–based approaches—and employs replicated-data causal inference to identify physically interpretable causal predictors, including vertical wind shear, mid-level potential vorticity, and surface humidity. Results: The selected causal features significantly improve model extrapolation to unseen storms, notably enhancing SHIPS+ skill for 24–72-hour forecasts. Several identified predictors have been validated operationally, delivering reliable intensity signals up to 120 hours in advance. This work establishes a novel, interpretable, robust, and physically consistent paradigm for tropical cyclone intensity prediction.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Knowledge Representation and Reasoning: Action, Change, and Causality

Application Category

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

Abstract

Improving statistical forecasts of Atlantic hurricane intensity is limited by complex nonlinear interactions and difficulty in identifying relevant predictors. Conventional methods prioritize correlation or fit, often overlooking confounding variables and limiting generalizability to unseen tropical storms. To address this, we leverage a multidata causal discovery framework with a replicated dataset based on Statistical Hurricane Intensity Prediction Scheme (SHIPS) using ERA5 meteorological reanalysis. We conduct multiple experiments to identify and select predictors causally linked to hurricane intensity changes. We train multiple linear regression models to compare causal feature selection with no selection, correlation, and random forest feature importance across five forecast lead times from 1 to 5 days (24 to 120 hours). Causal feature selection consistently outperforms on unseen test cases, especially for lead times shorter than 3 days. The causal features primarily include vertical shear, mid-tropospheric potential vorticity and surface moisture conditions, which are physically significant yet often underutilized in hurricane intensity predictions. Further, we build an extended predictor set (SHIPS+) by adding selected features to the standard SHIPS predictors. SHIPS+ yields increased short-term predictive skill at lead times of 24, 48, and 72 hours. Adding nonlinearity using multilayer perceptron further extends skill to longer lead times, despite our framework being purely regional and not requiring global forecast data. Operational SHIPS tests confirm that three of the six added causally discovered predictors improve forecasts, with the largest gains at longer lead times. Our results demonstrate that causal discovery improves hurricane intensity prediction and pave the way toward more empirical forecasts.

Problem

• Improving hurricane intensity forecasts using causal discovery
• Identifying causally linked predictors beyond correlation methods
• Enhancing prediction accuracy with physically significant atmospheric features

Innovation

• Multidata causal discovery framework identifies hurricane predictors
• Causal feature selection outperforms correlation-based methods
• Extended SHIPS+ predictor set increases short-term forecast skill