Tropical Cyclone Wind Field Diagnosis with Image-to-Image Models
Abstract
This study presents a novel application of image-to-image generative adversarial networks (GANs) to produce a complete tropical cyclone (TC) surface wind field diagnosis from a single frame of satellite infrared brightness temperature (IR). Because TC surface wind fields are infrequently observed, we train on two data sets. The first contains 3,323 image pairs of IR and wind fields from Weather Research and Forecasting Model (WRF) simulations (WRF2WRF). The second contains 11,886 real IR satellite images paired with parametric wind fields built on the extended best-track data set (IBTrACS) of TC intensities and wind radii (GOES2PARA). Both models, conditioned on storm intensity, motion, and latitude, transform single-channel IR into wind fields over 6∘ × 6∘ (WRF2WRF) and 7∘ × 7∘ (GOES2PARA) domains, at 256 × 256 pixel resolution. GOES2PARA captures outer wind field asymmetries, enabling diagnosis of 64, 50, and 34-knot wind radii (R64, R50, R34) in each quadrant. WRF2WRF also produces detailed rainbands and eyewall asymmetries. Pixel-wise RMSE against held out truth demonstrates realistic wind field generation with mean RMSE of 6.4 knots (GOES2PARA) and 7.2 knots (WRF2WRF). Compared against IBTrACS, WRF2WRF achieved mean absolute errors (MAE) of 14.8 nautical miles (NM) for radius of maximum wind (RMW) and 25.9 NM for R34. GOES2PARA achieved 10.0 NM MAE for RMW and 25.9 NM for R34. Against Synthetic Aperture Radar wind fields GOES2PARA yielded a lower mean pixel-wise error (RMSE 12.1 kt, MAE 9.5 kt) than WRF2WRF (RMSE 13.9 kt, MAE 11.3 kt). Unlike many existing operational surface wind speed algorithms, our image-to-image models provide updated wind fields with every new satellite image.
Related
- A Physics-Constrained Implicit Profile Network for Continuous Reconstruction of Tropical Cyclone Near-Surface Wind Profiles
- Downscaling of Tropical Cyclone Surface Wind Fields With a Hybrid Attention Transformer
- Skillful Seasonal Tropical Cyclone Prediction with AI Weather--Climate Models Forced by Dynamically Predicted SSTs
- Tibetan Plateau uplift increased western North Pacific tropical cyclone frequency
- Increase of tropical cyclone-induced coastal flooding due to historical sea level rise
BibTeX
@misc{ollier2026tropical,
title = {{Tropical Cyclone Wind Field Diagnosis with Image-to-Image Models}},
author = {Sarah J. Ollier and David S. Nolan and Brian D. McNoldy and Eduardo Siman},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.00038},
eprint = {2611.00038},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.00038},
}
RIS
TY - UNPB
TI - Tropical Cyclone Wind Field Diagnosis with Image-to-Image Models
AU - Sarah J. Ollier
AU - David S. Nolan
AU - Brian D. McNoldy
AU - Eduardo Siman
PY - 2026
DA - 2026/11/01
PB - arXiv
DO - 10.48550/arxiv.2611.00038
UR - https://arxiv.org/abs/2611.00038
AB - This study presents a novel application of image-to-image generative adversarial networks (GANs) to produce a complete tropical cyclone (TC) surface wind field diagnosis from a single frame of satellite infrared brightness temperature (IR). Because TC surface wind fields are infrequently observed, we train on two data sets. The first contains 3,323 image pairs of IR and wind fields from Weather Research and Forecasting Model (WRF) simulations (WRF2WRF). The second contains 11,886 real IR satellite images paired with parametric wind fields built on the extended best-track data set (IBTrACS) of TC intensities and wind radii (GOES2PARA). Both models, conditioned on storm intensity, motion, and latitude, transform single-channel IR into wind fields over 6^∘ × 6^∘ (WRF2WRF) and 7^∘ × 7^∘ (GOES2PARA) domains, at 256 × 256 pixel resolution. GOES2PARA captures outer wind field asymmetries, enabling diagnosis of 64, 50, and 34-knot wind radii (R64, R50, R34) in each quadrant. WRF2WRF also produces detailed rainbands and eyewall asymmetries. Pixel-wise RMSE against held out truth demonstrates realistic wind field generation with mean RMSE of 6.4 knots (GOES2PARA) and 7.2 knots (WRF2WRF). Compared against IBTrACS, WRF2WRF achieved mean absolute errors (MAE) of 14.8 nautical miles (NM) for radius of maximum wind (RMW) and 25.9 NM for R34. GOES2PARA achieved 10.0 NM MAE for RMW and 25.9 NM for R34. Against Synthetic Aperture Radar wind fields GOES2PARA yielded a lower mean pixel-wise error (RMSE 12.1 kt, MAE 9.5 kt) than WRF2WRF (RMSE 13.9 kt, MAE 11.3 kt). Unlike many existing operational surface wind speed algorithms, our image-to-image models provide updated wind fields with every new satellite image.
ER -