Global 3D Reconstruction of Clouds & Tropical Cyclones

TL;DR

This paper presents a novel machine learning framework that generates global 3D cloud maps from satellite data, enabling better analysis and forecasting of intense tropical cyclones.

Contribution

It introduces a pre-training--fine-tuning pipeline that reconstructs 3D cloud structures globally, including intense storms, from satellite imagery.

Findings

First global instantaneous 3D cloud maps of TCs

Accurate reconstruction of intense storm structures

Provides estimates even with missing satellite data

Abstract

Accurate forecasting of tropical cyclones (TCs) remains challenging due to limited satellite observations probing TC structure and difficulties in resolving cloud properties involved in TC intensification. Recent research has demonstrated the capabilities of machine learning methods for 3D cloud reconstruction from satellite observations. However, existing approaches have been restricted to regions where TCs are uncommon, and are poorly validated for intense storms. We introduce a new framework, based on a pre-training--fine-tuning pipeline, that learns from multiple satellites with global coverage to translate 2D satellite imagery into 3D cloud maps of relevant cloud properties. We apply our model to a custom-built TC dataset to evaluate performance in the most challenging and relevant conditions. We show that we can - for the first time - create global instantaneous 3D cloud maps and accurately reconstruct the 3D structure of intense storms. Our model not only extends available satellite observations but also provides estimates when observations are missing entirely. This is crucial for advancing our understanding of TC intensification and improving forecasts.