Geographic Transferability of Machine Learning Models for Short-Term Airport Fog Forecasting

TL;DR

This study demonstrates that physics-informed, coordinate-free machine learning models can effectively transfer fog forecasting capabilities across different geographic locations, outperforming traditional site-specific models.

Contribution

It introduces a coordinate-free feature set encoding thermodynamic and radiative processes, enabling geographic transferability of fog prediction models.

Findings

Achieved high AUC (0.923-0.947) across multiple sites and distances.

SHAP analysis shows physical features dominate, indicating learned transferable relationships.

Models outperform site-specific approaches in zero-shot transfer tests.

Abstract

Short-term forecasting of airport fog (visibility < 1.0 km) presents challenges in geographic generalization because many machine learning models rely on location-specific features and fail to transfer across sites. This study investigates whether fundamental thermodynamic and radiative processes can be encoded in a coordinate-free (location-independent) feature set to enable geographic transferability. A gradient boosting classifier (XGBoost) trained on Santiago, Chile (SCEL, 33S) data from 2002-2009 was evaluated on a 2010-2012 holdout set and under strict zero-shot tests at Puerto Montt (SCTE), San Francisco (KSFO), and London (EGLL). The model achieved AUC values of 0.923-0.947 across distances up to 11,650 km and different fog regimes (radiative, advective, marine). Consistent SHAP feature rankings show that visibility persistence, solar angle, and thermal gradients dominate predictions, suggesting the model learned transferable physical relationships rather than site-specific patterns. Results suggest that physics-informed, coordinate-free feature engineering can yield geographically transferable atmospheric forecasting tools.