Deep Learning Experiments for Tropical Cyclone Intensity Forecasts
Abstract
Reducing tropical cyclone (TC) intensity forecast errors is a challenging task that has interested the operational forecasting and research community for decades. To address this, we developed a deep learning (DL)-based multilayer perceptron (MLP) TC intensity prediction model. The model was trained using the global Statistical Hurricane Intensity Prediction Scheme (SHIPS) predictors to forecast the change in TC maximum wind speed for the Atlantic basin. In the first experiment, a 24-h forecast period was considered. To overcome sample size limitations, we adopted a leave one year out (LOYO) testing scheme, where a model is trained using data from all years except one and then evaluated on the year that is left out. When tested on 2010–18 operational data using the LOYO scheme, the MLP outperformed other statistical–dynamical models by 9%–20%. Additional independent tests in 2019 and 2020 were conducted to simulate real-time operational forecasts, where the MLP model again outperformed the statistical–dynamical models by 5%–22% and achieved comparable results as HWFI. The MLP model also correctly predicted more rapid intensification events than all the four operational TC intensity models compared. In the second experiment, we developed a lightweight MLP for 6-h intensity predictions. When coupled with a synthetic TC trackmore »
- Authors:
-
[1];
[1];
- Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
- Pacific Northwest National Lab. (PNNL), Richland, WA (United States); Mississippi State Univ., Mississippi State, MS (United States)
- Colorado State Univ., Fort Collins, CO (United States). Cooperative Institute for Research in the Atmosphere
- Publication Date:
- Research Org.:
- Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
- Sponsoring Org.:
- USDOE
- OSTI Identifier:
- 1818206
- Report Number(s):
- PNNL-SA-153486
Journal ID: ISSN 0882-8156
- Grant/Contract Number:
- AC05-76RL01830
- Resource Type:
- Accepted Manuscript
- Journal Name:
- Weather and Forecasting
- Additional Journal Information:
- Journal Volume: 36; Journal Issue: 4; Journal ID: ISSN 0882-8156
- Publisher:
- American Meteorological Society
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 58 GEOSCIENCES; 97 MATHEMATICS AND COMPUTING; Tropical cyclones; Forecast verification/skill; Forecasting techniques; Operational forecasting; Short-range prediction; Statistical forecasting; Deep learning; Neural network
Citation Formats
Xu, Wenwei, Balaguru, Karthik, August, Andrew, Lalo, Nicholas, Hodas, Nathan O., DeMaria, Mark, and Judi, David R. Deep Learning Experiments for Tropical Cyclone Intensity Forecasts. United States: N. p., 2021.
Web. doi:10.1175/WAF-D-20-0104.1.
Xu, Wenwei, Balaguru, Karthik, August, Andrew, Lalo, Nicholas, Hodas, Nathan O., DeMaria, Mark, & Judi, David R. Deep Learning Experiments for Tropical Cyclone Intensity Forecasts. United States. https://doi.org/10.1175/WAF-D-20-0104.1
Xu, Wenwei, Balaguru, Karthik, August, Andrew, Lalo, Nicholas, Hodas, Nathan O., DeMaria, Mark, and Judi, David R. Sun .
"Deep Learning Experiments for Tropical Cyclone Intensity Forecasts". United States. https://doi.org/10.1175/WAF-D-20-0104.1. https://www.osti.gov/servlets/purl/1818206.
@article{osti_1818206,
title = {Deep Learning Experiments for Tropical Cyclone Intensity Forecasts},
author = {Xu, Wenwei and Balaguru, Karthik and August, Andrew and Lalo, Nicholas and Hodas, Nathan O. and DeMaria, Mark and Judi, David R.},
abstractNote = {Reducing tropical cyclone (TC) intensity forecast errors is a challenging task that has interested the operational forecasting and research community for decades. To address this, we developed a deep learning (DL)-based multilayer perceptron (MLP) TC intensity prediction model. The model was trained using the global Statistical Hurricane Intensity Prediction Scheme (SHIPS) predictors to forecast the change in TC maximum wind speed for the Atlantic basin. In the first experiment, a 24-h forecast period was considered. To overcome sample size limitations, we adopted a leave one year out (LOYO) testing scheme, where a model is trained using data from all years except one and then evaluated on the year that is left out. When tested on 2010–18 operational data using the LOYO scheme, the MLP outperformed other statistical–dynamical models by 9%–20%. Additional independent tests in 2019 and 2020 were conducted to simulate real-time operational forecasts, where the MLP model again outperformed the statistical–dynamical models by 5%–22% and achieved comparable results as HWFI. The MLP model also correctly predicted more rapid intensification events than all the four operational TC intensity models compared. In the second experiment, we developed a lightweight MLP for 6-h intensity predictions. When coupled with a synthetic TC track model, the lightweight MLP generated realistic TC intensity distribution in the Atlantic basin. Therefore, the MLP-based approach has the potential to improve operational TC intensity forecasts, and will also be a viable option for generating synthetic TCs for climate studies.},
doi = {10.1175/WAF-D-20-0104.1},
journal = {Weather and Forecasting},
number = 4,
volume = 36,
place = {United States},
year = {Sun Aug 01 00:00:00 EDT 2021},
month = {Sun Aug 01 00:00:00 EDT 2021}
}
Search WorldCat to find libraries that may hold this journal