Hurricane Rapid Intensification Convolutional Neural Network: A Novel Deep Learning Model for Predicting Tropical Cyclone Rapid Intensification Using 20 Years of Satellite Rainfall Data
ISEF · 2023 Earth and Environmental Sciences Fourth Award
Overview
The rapid intensification (RI) of tropical cyclones (TCs) is one of the greatest challenges for forecasters. Improving the skill for RI forecasts has been a top priority of the National Hurricane Center (NHC). The Statistical Hurricane Intensity Prediction Scheme RI Index (SHIPS RII) is one of the most accurate models used operationally at NHC. The SHIPS RII uses predictors including climatology and persistence, environmental conditions, and infrared satellite information. One critical piece of information that is missing from SHIPS RII is the rainfall and structural features of TCs. In this study, I developed a novel Hurricane Rapid Intensification Convolutional Neural Network (HRICNN) model to predict the probability of future 24-h RI by using satellite rainfall images and SHIPS predictors. I designed the HRICNN model to detect specific TC structures such as the symmetry of the rainfall pattern associated with RI. A 20-year satellite rainfall dataset was obtained from the NASA Integrated Multi-satellite Retrievals for the Global Precipitation Measurement (GPM) mission (IMERG) product for TCs from the Atlantic basin. The HRICNN model was tested for 4 different radii of the IMERG data from the TC center and 300-km was selected. The model was trained using satellite images for TCs from 2000-2017 and tested using TCs from 2018-2019. My HRICNN model outperforms the SHIPS RII in terms of the Peirce Skill Score by 73% and Brier Skill Score by 13%. The HRICNN model’s probability of detection is 7% higher than the SHIPS RII, while the false alarm ratio is 3% lower.
Awards (2)
- Fourth Award of $500 $500
- National Geographic Society: Excellence in Geography and Geospatial Science Award
Competition history
- ISEF 2023
Resources
Related projects
ISEF · 2021
Predictive Modeling of Tropical Cyclone Rapid Intensification by Analyzing Convective Patterns with Convolutional Neural Networks, Year Four
ISEF · 2022
Predicting Future Tropical Cyclone Intensity Using a Convolutional Neural Network
ISEF · 2021
A Novel Deep Learning Model for Estimating Tropical Cyclone Intensity from Satellite Images
ISEF · 2024
An Artificial Intelligence Tool for Intensity Estimation of Global Tropical Cyclones
Closest projects by meaning, across every fair and year in the corpus.
Source: Regeneron International Science and Engineering Fair