Machine Learning Algorithms for Tropical Cyclone Center Fixing and Eye Detection

The production of a storm center estimate is usually one of the first steps in performing a forecast for a tropical cyclone and has an impact on all downstream forecasts. Determining the onset of eye formation is also very important for intensity forecasts, and for center estimation. Currently, most existing operational eye detection and center-fix methods are subjective and little investigation has been made into the use of objective techniques. Using techniques from the fields of machine learning and computer vision, three applications are under development to investigate the use of objective techniques to perform eye detection and center fixing. The preliminary results for two center fixing algorithms, one using Advanced Microwave Sounding Unit (AMSU) and Advanced Technology Microwave Sounder (ATMS) data with Quadratic Discriminant Analysis (QDA) and one using GOES IR data with the Circular Hough Transform (CHT), are presented here. The algorithm using QDA showed a 10% improvement over a baseline real-time estimate provided by the National Hurricane Center (NHC). The algorithm using the CHT showed no improvement compared to the NHC baseline estimate but may improve with future work. Additionally, work completed on an algorithm to perform eye detection using GOES IR data with QDA and other machine learning algorithms will be presented.

Disclaimer: The views, opinions, and findings contained in this article are those of the authors and should not be construed as an official National Oceanic and Atmospheric Administration (NOAA) or U.S. Government position, policy, or decision.