Center Fixing Tropical Depressions and Tropical Storms Using Machine Learning - Nighttime Visible Imagery

Aug 5, 2026

The first step in tropical cyclone (TC) retrieval algorithms is determining the storm’s central position, yet satellite-based centers can be obscured during intensifying and decaying phases by thick clouds or overlying cirrus. Microwave imagery helps but often suffers discontinuous coverage from polar-orbiting sensors across the TC life cycle. This study evaluates machine learning-derived Nighttime Visible Imagery (NVI) to improve center fix accuracy for tropical depression (TD)- and tropical storm (TS)-strength systems in darkness. NVI is generated from Himawari channels 7, 11, 13, and 15 with brightness temperature differences mapped to lunar reflectance comparable to DNB measurements.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages72

Reading Time3 h 1 min

AccessFree to download

Abstract

Background and Literature Review

Tropical Cyclone Satellite Scenes

Objective Center Fixing

Radiative Transfer

Machine Learning

Methodology

Air Force Institute of Technology AFIT Scholar | Theses and Dissertations | Student Graduate Works | | --- | --- | | 3-2024
Center Fixing Tropical Depressions and Tropical Storms Using Machine Learning-Nighttime Visible Imagery
Nathan K. Stanford
Follow this and additional works at: [https://scholar.afit.edu/etd](https://scholar.afit.edu/etd)
Part of the Meteorology Commons | | Recommended Citation Stanford, Nathan K., "Center Fixing Tropical Depressions and Tropical Storms Using Machine LearningNighttime Visible Imagery" (2024) . Theses and Dissertations. 7787. [https://scholar.afit.edu/etd/7787](https://scholar.afit.edu/etd/7787) This Thesis is brought to you for free and open access by the Student Graduate Works at AFIT Scholar. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of AFIT Scholar. For more information, [please contact](please contact AFIT.ENWL.Repository@us.af.mil)[ AFIT.ENWL.Repository@us.af.mil](please contact AFIT.ENWL.Repository@us.af.mil). CENTER FIXING TROPICAL DEPRESSIONSAND TROPICAL STORMS USING MACHINE LEARNING-NIGHTTIME VISIBLE IMAGERY THESIS Nathan K. Stanford, Captain, USAF AFIT-METG-THESIS DEPARTMENT OF THE AIR FORCE AIR UNIVERSITY AIR FORCE INSTITUTE OF TECHNOLOGY Wright-Patterson Air Force Base, Ohio APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED The views expressed in this document are those of the author and do not reflect the official policy or position of the United States Air Force, the United States Department of Defense or the United States Government. This material is declared a work of the U.S. Government and is not subject to copyright protection in the United States. AFIT-METG-THESIS CENTER FIXING TROPICAL DEPRESSIONS AND TROPICAL STORMS USING MACHINE LEARNING-NIGHTTIME VISIBLE IMAGERY THESIS Presented to the Faculty Department of Engineering Physics Graduate School of Engineering and Management Air Force Institute of Technology Air University Air Education and Training Command in Partial Fulfillment of the Requirements for the Degree of Master of Science in Meteorology Nathan K. Stanford, B.S.M. , M.S.P.M. Captain, USAF March 4, 2024 APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED AFIT-METG-THESIS CENTER FIXING TROPICAL DEPRESSIONS AND TROPICAL STORMS USING MACHINE LEARNING-NIGHTTIME VISIBLE IMAGERY THESIS Nathan K. Stanford, B.S.M. , M.S.P.M. Captain, USAF Committee Membership: Chandra M. Pasillas, Ph.D Chair Anthony J. Wimmers, Ph.D Member Matthew F. Horan, Ph.D Member AFIT-METG-THESIS Abstract The first step in most tropical cyclone (TC)-retrieval algorithms is determining the storm’s central position. When viewed in satellite imagery, mature TC centers are often often highlighted by a distinct eye and curved band pattern; however, in intensifying and decaying storms, the center is often obscured by thick clouds or overlying cirrus. Generally, microwave imagery addresses this challenge; however, microwave sensors aboard polar orbiting satellites provide discontinuous coverage throughout the TC life cycle; therefore, analysts rely on geostationary satellite imagery for continuous storm coverage. This study assesses the benefits of incorporating machine learningderived nighttime visible imagery to improve center fix accuracy in intensifying and decaying tropical depression (TD)-and tropical storm (TS)-strength TCs during periods of darkness. Machine Learning-Nighttime Visible Imagery (NVI) incorporates Himawari Channels 7 (3.9µm), 11 (8.6µm), 13 (10.4µm), and 15 (12.4µm), as well as their brightness temperature differences, to create lunar reflectance comparable to the lunar reflectance values derived from Day/Night Band (DNB) measured radiances. The study is divided into two parts: the first, an objective analysis using the Automated Rotational Center Hurricane Eye Retrieval (ARCHER-2) algorithm, and the second, a subjective imag

FAQs

Why is center fixing difficult for tropical cyclones during intensifying and decaying stages?

Mature centers may be obscured by thick clouds or cirrus, making the central position hard to identify in satellite imagery. The study highlights this limitation and the need for better nighttime-aware approaches.

What is Nighttime Visible Imagery (NVI) in this thesis?

NVI is machine learning-derived imagery that uses Himawari channels 7, 11, 13, and 15 plus brightness temperature differences to produce lunar reflectance comparable to Day/Night Band (DNB) measurements.

How does the proposed approach improve center fix accuracy?

In the objective analysis using ARCHER-2, NVI improved center fix assessments by 21.9 km versus SWIR and by 64.5 km versus LWIR. In the subjective analysis, analysts favored SWIR by 30.9 km based on 7 Western Pacific TC image sets.

Aug 5, 2026

Upload by Evangeline