# Global 3D Reconstruction of Clouds & Tropical Cyclones

arXiv: https://arxiv.org/abs/2511.04773 · HTML: https://mlsanity.com/paper/2511.04773/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 33 references that are in the map.
- First posted on arXiv 6 Nov 2025.

## Builds on (most cited first, 8 of 8)

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [SatMAE: Pre-training Transformers for Temporal and Multi-Spectral Satellite Imagery](https://mlsanity.com/paper/2207.08051/) (2022)
- [ResUNet-a: a deep learning framework for semantic segmentation of remotely sensed data](https://mlsanity.com/paper/1904.00592/) (2019)
- [Swin MAE: Masked Autoencoders for Small Datasets](https://mlsanity.com/paper/2212.13805/) (2022)
- [3D Cloud reconstruction through geospatially-aware Masked Autoencoders](https://mlsanity.com/paper/2501.02035/) (2025)
- [IceCloudNet: 3D reconstruction of cloud ice from Meteosat SEVIRI](https://mlsanity.com/paper/2410.04135/) (2024)

## Built on by (most cited first, 1 of 1)

- [CERBERUS: A Three-Headed Decoder for Vertical Cloud Profiles](https://mlsanity.com/paper/2604.08772/) (2026)

## Abstract (arXiv)

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.