Physics-Informed Diffusion Model for Generating Synthetic Extreme Rare Weather Events Data

TL;DR

This paper introduces a physics-informed diffusion model conditioned on atmospheric parameters to generate realistic synthetic satellite imagery of extreme weather events, addressing data scarcity and class imbalance in tropical cyclone detection.

Contribution

The novel physics-informed diffusion model preserves physical consistency and spatial autocorrelation in synthetic weather data, improving data augmentation for rare extreme events.

Findings

Successfully generates realistic multi-spectral satellite imagery of extreme weather events.

Effectively mitigates data imbalance, especially for rare Category 4-equivalent events.

Achieves an average Log-Spectral Distance of 4.5dB, indicating high quality of generated data.

Abstract

Data scarcity is a primary obstacle in developing robust Machine Learning (ML) models for detecting rapidly intensifying tropical cyclones. Traditional data augmentation techniques (rotation, flipping, brightness adjustment) fail to preserve the physical consistency and high-intensity gradients characteristic of rare Category 4-equivalent events, which constitute only 0.14\% of our dataset (202 of 140,514 samples). We propose a physics-informed diffusion model based on the Context-UNet architecture to generate synthetic, multi-spectral satellite imagery of extreme weather events. Our model is conditioned on critical atmospheric parameters such as average wind speed, type of Ocean and stage of development (early, mature, late etc) -- the known drivers of rapid intensification. Using a controlled pre-generated noise sampling strategy and mixed-precision training, we generated $16\times16$ wind-field samples that are cropped from multi-spectral satellite imagery which preserve realistic spatial autocorrelation and physical consistency. Results demonstrate that our model successfully learns discriminative features across ten distinct context classes, effectively mitigating the data bottleneck. Specifically, we address the extreme class imbalance in our dataset, where Class 4 (Ocean 2, early stage with average wind speed 50kn hurricane) contains only 202 samples compared to 79,768 samples in Class 0. This generative framework provides a scalable solution for augmenting training datasets for operational weather detection algorithms. The average Results yield an average Log-Spectral Distance (LSD) of 4.5dB, demonstrating a scalable framework for enhancing operational weather detection algorithms.