Exploring Tropical Cyclone Structure and Evolution with AI-based Synthetic Passive Microwave Data

Satellite observations are essential for monitoring tropical cyclone formation and evolution, gathering information about tropical cyclone structure and behavior, and analyzing tropical cyclone position and intensity. Geostationary satellites provide valuable visible and infrared window images of tropical cyclones with broad spatial coverage and high temporal frequency; however, these channels are primarily limited to observing at cloud top features. Passive microwave imagery can see through the clouds, providing valuable information about tropical cyclone structure and evolution; however, these instruments are limited to at most a handful of overpasses per day due to the intersection of their low-Earth orbits with tropical cyclone positions. Recently, researchers at CIRA and NESDIS have used machine learning to construct a synthetic passive microwave (SPM) dataset. The SPM dataset ingests the full suite of visible and infrared imagery from geostationary satellites into an ensemble of fully-connected and convolutional neural network configurations that output simulated 89 GHz passive microwave imagery at high temporal frequency. This SPM dataset, consisting of thousands of images of global tropical cyclones observed over the past five years, forms the basis of an exploration of various aspects of tropical cyclone structure and evolution at higher temporal frequencies. The analysis is focused on statistical and artificial intelligence methods, exploring the possibilities of using this machine learning-generated dataset to provide valuable physical intuition regarding the structural evolution of tropical cyclones across the globe.

Disclaimer: The scientific results and conclusions, as well as any views or opinions expressed herein, are those of the author(s) and do not necessarily reflect those of NOAA or the Department of Commerce.