Calibration of Typhoon Track Forecasts Based on Deep Learning Methods

An accurate forecast of typhoon tracks is crucial for disaster warning and mitigation. However, existing numerical weather prediction models, such as the Weather Research and Forecasting (WRF) model, still exhibit significant errors in track forecasts. This study aims to improve forecast accuracy by...

Full description

Saved in:
Bibliographic Details
Published inAtmosphere Vol. 15; no. 9; p. 1125
Main Authors Tao, Chengchen, Wang, Zhizu, Tian, Yilun, Han, Yaoyao, Wang, Keke, Li, Qiang, Zuo, Juncheng
Format Journal Article
LanguageEnglish
Published Basel MDPI AG 01.09.2024
Subjects
Accuracy
Artificial intelligence
BiLSTM
Climate models
Comparative analysis
ConvLSTM
Cyclones
Datasets
Deep learning
Efficiency
Emergency warning programs
Environmental aspects
error decomposition
Error reduction
Extreme values
Forecast accuracy
Forecasts and trends
Hurricanes
Kalman filters
Long short-term memory
Machine learning
Mitigation
Neural networks
Numerical weather forecasting
Performance evaluation
Prediction models
Radiation
Typhoon forecasting
typhoon track forecast
Typhoon tracks
Typhoons
Weather
Weather forecasting
Wind
WRF model
China
Online AccessGet full text

Cover

More Information
Summary:An accurate forecast of typhoon tracks is crucial for disaster warning and mitigation. However, existing numerical weather prediction models, such as the Weather Research and Forecasting (WRF) model, still exhibit significant errors in track forecasts. This study aims to improve forecast accuracy by correcting WRF-forecasted tracks using deep learning models, including Bidirectional Long Short-Term Memory (BiLSTM) + Convolutional Long Short-Term Memory (ConvLSTM) + Wide and Deep Learning (WDL), BiLSTM + Convolutional Gated Recurrent Unit (ConvGRU) + WDL, and BiLSTM + ConvLSTM + Extreme Deep Factorization Machine (xDeepFM), with a comparison to the Kalman Filter. The results demonstrate that the BiLSTM + ConvLSTM + WDL model reduces the 72 h track prediction error (TPE) from 255.18 km to 159.23 km, representing a 37.6% improvement over the original WRF model, and exhibits significant advantages across all evaluation metrics, particularly in key indicators such as Bias2, Mean Squared Error (MSE), and Sequence. The decomposition of MSE further validates the importance of the BiLSTM, ConvLSTM, WDL, and Temporal Normalization (TN) layers in enhancing the model’s spatio-temporal feature-capturing ability.
ISSN:2073-4433
2073-4433
DOI:10.3390/atmos15091125