Rapid intensification forecast of tropical cyclones based on machine learning

doi:10.3969/j.issn.1001-909X.2024.03.008

Abstract

Abstract:

Extremely deep convective clouds are a precursor to the rapid intensification (RI) of tropical cyclones (TCs). To predict the RI of TCs in the western North Pacific (WNP), a machine learning (ML) model using data related to extremely deep convective clouds was developed. The ML model integrates the Statistical Hurricane Intensity Prediction Scheme-Rapid Intensification Index (SHIPS-RII) data and the coverage area of extremely deep convective clouds within a 300 km radius of the center of the TC. Based on data from 2011 to 2019, the model forecasted RI events that increased by more than 30 kn and 35 kn within 24 h. Compared to models using only SHIPS-RII data, this ML model showed an improvement of 5.66% and 9.58% in the Peirce Skill Score (PSS), and a relative increase of 8.41% and 8.55% in the probability of detection (POD). This model is used to forecast typical typhoon Dujuan (2015), and the results show that the model integrating the coverage area of extremely deep convective clouds has advantages in RI predictions, which is mainly reflected in RI forecasts when the initial intensity is strong. The model has great application potential for forecasting strong typhoons.

Key words: western North Pacific, tropical cyclones, rapid intensification, infrared brightness temperature (IR BT), extremely deep convective clouds, machine learning, typhoon Dujuan(2015), initeal intensity of TC

CLC Number:

P732.4

LUO Tong, HONG Jiacheng. Rapid intensification forecast of tropical cyclones based on machine learning[J]. Journal of Marine Sciences, 2024, 42(3): 99-107.

Figures/Tables 9

Tab.1 Predictive factors utilized in SHIPS-RII

序号	预报因子	定义
1	PER	前12 h TC强度变化
2	VMX0	最大持续风速(t=0 h)
3	POT	TC潜在强度(TC当前强度-最大可能强度)
4	OHC	海洋热含量
5	SHRD	滤波后500 km半径范围内850 hPa与200 hPa之间的水平风垂直切变
6	D200	1 000 km半径范围内200 hPa散度
7	TPW	500 km半径范围内以上切方向为中心的90°方位象限内总降水量小于45 mm的区域所占百分比
8	PC2	440 km半径范围内GOES-IR图像的第二主成分(t=0 h)
9	SDBT	50~200 km半径范围内GOES BT的标准差

Fig.1 Flowchart of model training and testing process

Fig.2 Composite image showing the variation in average brightness temperature of extremely deep convective clouds within a 300 km radius of the Western North Pacific (WNP) TC center from 2000 to 2019, as a function of TC intensity change rate

Fig.3 Trends of coverage area ratio and average brightness temperature of extremely deep convective clouds with TC intensity change rate within a 300 km radius of the TC center in the WNP from 2000 to 2019

Fig.4 Comparison of prediction capabilities of the SHIPS-RII model and the ML.Base model for TC rapid intensification in WNP from 2011 to 2019

Fig.5 Comparison of prediction capabilities of the ML.Base model and the ML.DC200 model for TC rapid intensification in WNP from 2011 to 2019

Fig.6 Comparison of prediction capabilities of the ML.Base model and the ML.DC300 model for TC rapid intensification in WNP from 2011 to 2019

Fig.7 The track during typhoon Dujuan’s intensification in 2015 (RI represents where rapid intensification begins, blue line indicates rapid intensification and red line indicates non-rapid intensification.)

Fig.8 Variation of coverage area of extremely deep convective clouds within a 300 km radius of the TC center and TC intensity during the life cycle of typical typhoon Dujuan (RI represents where rapid intensification begins.)

References 23

[1]	张娇艳, 吴立广, 张强. 全球变暖背景下我国热带气旋灾害趋势分析[J]. 热带气象学报, 2011, 27(4):442-454.
	ZHANG J Y, WU L G, ZHANG Q. Tropical cyclone damages in China under the background of global warming[J]. Journal of Tropical Meteorology, 2011, 27(4): 442-454.
[2]	黄昌兴, 周国良, 郑磊, 等. 登陆我国台风的时空分布特征及其影响[J]. 水文, 2014, 34(6):81-85.
	HUANG C X, ZHOU G L, ZHENG L, et al. Spatial and temporal distribution characteristics of landing typhoons in China and their influence[J]. Journal of China Hydrology, 2014, 34(6): 81-85.
[3]	CANGIALOSI J P, BLAKE E, DEMARIA M, et al. Recent progress in tropical cyclone intensity forecasting at the national hurricane center[J]. Weather and Forecasting, 2020, 35(5): 1913-1922.
[4]	TRABING B C, BELL M M. Understanding error distributions of hurricane intensity forecasts during rapid intensity changes[J]. Weather and Forecasting, 2020, 35(6): 2219-2234.
[5]	KAPLAN J, DEMARIA M. Large-scale characteristics of rapidly intensifying tropical cyclones in the North Atlantic Basin[J]. Weather and Forecasting, 2003, 18(6): 1093-1108.
[6]	HENDRICKS E A, PENG M S, FU B, et al. Quantifying environmental control on tropical cyclone intensity change[J]. Monthly Weather Review, 2010, 138(8): 3243-3271.
[7]	CHEN H, ZHANG D L. On the rapid intensification of hurricane Wilma (2005). part II: Convective bursts and the upper-level warm core[J]. Journal of the Atmospheric Sciences, 2013, 70(1): 146-162.
[8]	CHEN X M, WANG Y Q, FANG J, et al. A numerical study on rapid intensification of typhoon Vicente (2012) in the South China Sea. Part II: Roles of inner-core processes[J]. Journal of the Atmospheric Sciences, 2018, 75(1): 235-255.
[9]	HAZELTON A T, HART R E, ROGERS R F. Analyzing simulated convective bursts in two Atlantic hurricanes. Part II: Intensity change due to bursts[J]. Monthly Weather Review, 2017, 145(8): 3095-3117.
[10]	JIANG H Y. The relationship between tropical cyclone intensity change and the strength of inner-core convection[J]. Monthly Weather Review, 2012, 140(4): 1164-1176.
[11]	MOLINARI J, VOLLARO D. Rapid intensification of a sheared tropical storm[J]. Monthly Weather Review, 2010, 138(10): 3869-3885.
[12]	ROGERS R F, REASOR P D, ZHANG J A. Multiscale structure and evolution of hurricane Earl (2010) during rapid intensification[J]. Monthly Weather Review, 2015, 143(2): 536-562.
[13]	SHI D L, CHEN G H, XIE X R. Revisiting the relationship between tropical cyclone rapid intensification and the distribution of inner-core precipitation[J]. Geophysical Research Letters, 2023, 50(17): e2023GL104810.
[14]	GAO S, ZHANG W, LIU J, et al. Improvements in typhoon intensity change classification by incorporating an ocean coupling potential intensity index into decision trees[J]. Weather and Forecasting, 2016, 31(1): 95-106.
[15]	GENG H T, SHI D W, ZHANG W, et al. A prediction scheme for the frequency of summer tropical cyclone landfalling over China based on data mining methods[J]. Meteorological Applications, 2016, 23(4): 587-593.
[16]	MERCER A, GRIMES A. Diagnosing tropical cyclone rapid intensification using Kernel methods and reanalysis datasets[J]. Procedia Computer Science, 2015, 61: 422-427.
[17]	SU H, WU L T, JIANG J H, et al. Applying satellite observations of tropical cyclone internal structures to rapid intensification forecast with machine learning[J]. Geophysical Research Letters, 2020, 47(17): e2020GL089102.
[18]	RUAN Z X, WU Q Y. Precipitation, convective clouds, and their connections with tropical cyclone intensity and intensity change[J]. Geophysical Research Letters, 2018, 45(2): 1098-1105.
[19]	WU Q Y, HONG J C, RUAN Z X. Diurnal variations in tropical cyclone intensification[J]. Geophysical Research Letters, 2020, 47(23): e2020GL090397.
[20]	JANOWIAK J E, JOYCE R J, YAROSH Y. A real-time global half-hourly pixel-resolution infrared dataset and its applications[J]. Bulletin of the American Meteorological Society, 2001, 82(2): 205-217.
[21]	WU Q Y, RUAN Z X. Rapid contraction of the radius of maximum tangential wind and rapid intensification of a tropical cyclone[J]. Journal of Geophysical Research: Atmospheres, 2021, 126(3): e2020JD033681.
[22]	KAPLAN J, ROZOFF C M, DEMARIA M, et al. Evaluating environmental impacts on tropical cyclone rapid intensifi-cation predictability utilizing statistical models[J]. Weather and Forecasting, 2015, 30(5): 1374-1396.
[23]	KAPLAN J, DEMARIA M, KNAFF J A. A revised tropical cyclone rapid intensification index for the Atlantic and eastern North Pacific basins[J]. Weather and Forecasting, 2010, 25(1): 220-241.