Western North Pacific tropical cyclone track forecasts by a machine learning model
- PDF / 3,091,723 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 10 Downloads / 178 Views
ORIGINAL PAPER
Western North Pacific tropical cyclone track forecasts by a machine learning model Jinkai Tan1,2,3
•
Sheng Chen1,2 • Jun Wang3,4
Accepted: 3 November 2020 Ó Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract An ensemble machine learning model for tropical cyclone (TC) track forecasts over the Western North Pacific was developed and evaluated in this study. First, we investigated predictors including TC climatology and persistence factors which were extracted from TC best-track dataset and storm’s surrounding atmospheric conditions which were extracted from ERA-Interim reanalysis. Then, we built a Gradient Boosting Decision Tree (GBDT) nonlinear model for TC track forecasts, in which 30-year data was used. Finally, using tenfold cross-validation method, the GBDT model was compared with a frequently used technique: climatology and persistence (CLIPER) model. The experimental results show that the GBDT model performs well in three forecast times (24 h, 48 h, and 72 h) with relatively small forecast error of 138, 264, and 363.5 km, respectively. The model obtains excellent TC moving direction aspects. However, the model is still insufficient to produce aspects of storm acceleration and deceleration, with mean moving velocity sensitivities all less than 60%. Nevertheless, the model obtains much more robust and accurate TC tracks relative to CLIPER model, where the forecast skills are 17.5%, 26.3%, and 32.1% at three forecast times, respectively. The presented study demonstrates that the GBDT model could provide reliable evidence and guidance for operational TC track forecasts. Keywords Tropical cyclone Tracks Forecast Machine Learning
1 Introduction
& Jun Wang [email protected] Jinkai Tan [email protected] Sheng Chen [email protected] 1
School of Atmospheric Sciences, and Key Laboratory of Tropical Atmosphere-Ocean System (Ministry of Education), Sun Yat-sen University, Zhuhai 519000, China
2
Guangdong Province Key Laboratory for Climate Change and Natural Disaster Studies, Sun Yat-sen University, Zhuhai 519000, China
3
School of Geographic Sciences, East China Normal University, Shanghai 200241, China
4
Key Laboratory of Geographic Information Science (Ministry of Education), East China Normal University, Shanghai 200241, China
TC track forecast is one of the key components in TC risk assessment studies (Nakamura et al. 2015, 2017). Over the past few decades, many researchers have studied TC track forecasts by using a variety of measures, of which three common types are: dynamic forecast, numerical weather prediction (NWP) systems, and statistical forecast technique (Roy and Kovorda´nyi 2012). Dynamic forecast mainly describes the weather features and TC ambient environments, and imitates a vortex-like cyclone from a large-scale meteorological field to estimate TC motions (Holland 1993; Goerss and Jeffries 1994; Elsberry 1995). However, it is both computational resource and time consuming in practice. NWP is a multi-model ensemble system consisting of si
Data Loading...
