Hyeong-Seog Kim
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Hyeong-Seog Kim has a total of 100 co-authors. Here are the top co-authors by number of shared publications:
- Chang-Hoi Ho
Department of Climate and Energy Systems Engineering, Ewha Womans University, Seoul, Republic of Korea
13 shared publications Chang-Hoi Ho
Department of Climate and Energy Systems Engineering, Ewha Womans University, Seodaemun-gu, Seoul, South Korea
6 shared publications- Doo-Sun R. Park
Kyungpook National University
5 shared publications Joo-Hong Kim
Division of Climate Change, Korea Polar Research Institute, Incheon, Republic of Korea
5 shared publicationsDoo-Sun R. Park
Korea Adaptation Center for Climate Change, Korea Environment Institute, Sejong, South Korea
4 shared publications- Dong-Hyun Cha
Department of Civil, Urban, Earth and Environmental Engineering, Ulsan National Institute of Science and Technology, Ulsan, 44919, Republic of Korea
3 shared publications
Latest Publications
- Open Access
- 2 November 2026
- Published by Copernicus GmbH
- Vol. 2026, 1-30
Comment on egusphere-2026-3285
Accurate detection of tropical cyclones (TCs) from gridded climate model data is essential for evaluating model performance and projecting future TC activity. Conventional detection schemes rely on environmental variable thresholds that are frequently tuned to specific basins or models, making them inherently subjective. Conversely, detection schemes based on universal thresholds often fail to capture regional characteristics. While recent machine learning (ML) approaches provide objective data-driven thresholds, they generally require a great number of variables compared to conventional methods and suffer from high false alarm rates (FAR). Here, we introduce an objective hybrid tracker and classifier for tropical cyclones (TC2), an algorithm combining traditional and ML techniques to establish objective thresholds and minimize FAR. Using an ensemble of six classifiers based on three ML algorithms and two reanalyses, TC2 avoids dependency on specific ML algorithms or datasets. TC2 was trained and its hyperparameters were optimized using two reanalysis datasets over 1998–2017 period, while its performance was evaluated on their respective internal test sets and the independent NCEP FNL dataset from 2018 to 2024. Evaluated against the NCEP FNL dataset, TC2 outperforms the existing algorithms, i.e., TempestExtremes and OWZP, achieving higher F1 score (83.6 %) and critical success index (71.8 %), while significantly lowering FAR (14.3 %) and maintaining a comparable hit rate (81.5 %). TC2 also better reproduces TC count and seasonal variability over each basin. In CMIP6 evaluations, TC2 successfully captures the overall characteristics of TC activity. Under the SSP2-4.5 scenario, projected spatial changes in TC genesis frequency detected by TC2 are largely consistent with those of the dynamical genesis potential index, suggesting that TC2 identifies physically coherent systems governed by large-scale dynamic environments. Utilizing only a limited set of commonly available variables, including minimum sea level pressure, low-level relative vorticity, upper- and low-level wind speeds, and an upper-level warm core, TC2 provides an effective and robust framework for TC detection in gridded climate datasets.
- 2 November 2026
- Published by Copernicus GmbH
- Vol. 2026, 1-30
Open AccessComment on egusphere-2026-3285
Accurate detection of tropical cyclones (TCs) from gridded climate model data is essential for evaluating model performance and projecting future TC activity. Conventional detection schemes rely on environmental variable thresholds that are frequently tuned to specific basins or models, making them inherently subjective. Conversely, detection schemes based on universal thresholds often fail to capture regional characteristics. While recent machine learning (ML) approaches provide objective data-driven thresholds, they generally require a great number of variables compared to conventional methods and suffer from high false alarm rates (FAR). Here, we introduce an objective hybrid tracker and classifier for tropical cyclones (TC2), an algorithm combining traditional and ML techniques to establish objective thresholds and minimize FAR. Using an ensemble of six classifiers based on three ML algorithms and two reanalyses, TC2 avoids dependency on specific ML algorithms or datasets. TC2 was trained and its hyperparameters were optimized using two reanalysis datasets over 1998–2017 period, while its performance was evaluated on their respective internal test sets and the independent NCEP FNL dataset from 2018 to 2024. Evaluated against the NCEP FNL dataset, TC2 outperforms the existing algorithms, i.e., TempestExtremes and OWZP, achieving higher F1 score (83.6 %) and critical success index (71.8 %), while significantly lowering FAR (14.3 %) and maintaining a comparable hit rate (81.5 %). TC2 also better reproduces TC count and seasonal variability over each basin. In CMIP6 evaluations, TC2 successfully captures the overall characteristics of TC activity. Under the SSP2-4.5 scenario, projected spatial changes in TC genesis frequency detected by TC2 are largely consistent with those of the dynamical genesis potential index, suggesting that TC2 identifies physically coherent systems governed by large-scale dynamic environments. Utilizing only a limited set of commonly available variables, including minimum sea level pressure, low-level relative vorticity, upper- and low-level wind speeds, and an upper-level warm core, TC2 provides an effective and robust framework for TC detection in gridded climate datasets.
- Open Access
- 9 July 2026
- Published by Copernicus GmbH in EGUsphere
- Vol. 2026, 1-30
TC 2 ver. 1.0: An Objective Hybrid Tracker and Classifier for Tropical Cyclones version 1.0
Accurate detection of tropical cyclones (TCs) from gridded climate model data is essential for evaluating model performance and projecting future TC activity. Conventional detection schemes rely on environmental variable thresholds that are frequently tuned to specific basins or models, making them inherently subjective. Conversely, detection schemes based on universal thresholds often fail to capture regional characteristics. While recent machine learning (ML) approaches provide objective data-driven thresholds, they generally require a great number of variables compared to conventional methods and suffer from high false alarm rates (FAR). Here, we introduce an objective hybrid tracker and classifier for tropical cyclones (TC2), an algorithm combining traditional and ML techniques to establish objective thresholds and minimize FAR. Using an ensemble of six classifiers based on three ML algorithms and two reanalyses, TC2 avoids dependency on specific ML algorithms or datasets. TC2 was trained and its hyperparameters were optimized using two reanalysis datasets over 1998–2017 period, while its performance was evaluated on their respective internal test sets and the independent NCEP FNL dataset from 2018 to 2024. Evaluated against the NCEP FNL dataset, TC2 outperforms the existing algorithms, i.e., TempestExtremes and OWZP, achieving higher F1 score (83.6 %) and critical success index (71.8 %), while significantly lowering FAR (14.3 %) and maintaining a comparable hit rate (81.5 %). TC2 also better reproduces TC count and seasonal variability over each basin. In CMIP6 evaluations, TC2 successfully captures the overall characteristics of TC activity. Under the SSP2-4.5 scenario, projected spatial changes in TC genesis frequency detected by TC2 are largely consistent with those of the dynamical genesis potential index, suggesting that TC2 identifies physically coherent systems governed by large-scale dynamic environments. Utilizing only a limited set of commonly available variables, including minimum sea level pressure, low-level relative vorticity, upper- and low-level wind speeds, and an upper-level warm core, TC2 provides an effective and robust framework for TC detection in gridded climate datasets.
- 9 July 2026
- Published by Copernicus GmbH in EGUsphere
- Vol. 2026, 1-30
Open AccessTC 2 ver. 1.0: An Objective Hybrid Tracker and Classifier for Tropical Cyclones version 1.0
Accurate detection of tropical cyclones (TCs) from gridded climate model data is essential for evaluating model performance and projecting future TC activity. Conventional detection schemes rely on environmental variable thresholds that are frequently tuned to specific basins or models, making them inherently subjective. Conversely, detection schemes based on universal thresholds often fail to capture regional characteristics. While recent machine learning (ML) approaches provide objective data-driven thresholds, they generally require a great number of variables compared to conventional methods and suffer from high false alarm rates (FAR). Here, we introduce an objective hybrid tracker and classifier for tropical cyclones (TC2), an algorithm combining traditional and ML techniques to establish objective thresholds and minimize FAR. Using an ensemble of six classifiers based on three ML algorithms and two reanalyses, TC2 avoids dependency on specific ML algorithms or datasets. TC2 was trained and its hyperparameters were optimized using two reanalysis datasets over 1998–2017 period, while its performance was evaluated on their respective internal test sets and the independent NCEP FNL dataset from 2018 to 2024. Evaluated against the NCEP FNL dataset, TC2 outperforms the existing algorithms, i.e., TempestExtremes and OWZP, achieving higher F1 score (83.6 %) and critical success index (71.8 %), while significantly lowering FAR (14.3 %) and maintaining a comparable hit rate (81.5 %). TC2 also better reproduces TC count and seasonal variability over each basin. In CMIP6 evaluations, TC2 successfully captures the overall characteristics of TC activity. Under the SSP2-4.5 scenario, projected spatial changes in TC genesis frequency detected by TC2 are largely consistent with those of the dynamical genesis potential index, suggesting that TC2 identifies physically coherent systems governed by large-scale dynamic environments. Utilizing only a limited set of commonly available variables, including minimum sea level pressure, low-level relative vorticity, upper- and low-level wind speeds, and an upper-level warm core, TC2 provides an effective and robust framework for TC detection in gridded climate datasets.
- Open Access
- 30 April 2026
- Published by IOP Publishing in Environmental Research Letters
- Vol. 21 (10) , 104013
Modest ocean warming difference, explosive intensification: Marine heatwave control of Typhoon Hagibis (2019)
Typhoon Hagibis (2019) explosively intensified by 100 kt in 24 hours, greatly exceeding the conventional rapid intensification (RI) threshold of ~30 kt per 24 h. This study investigates how marine heatwave (MHW)–driven ocean surface and subsurface structures influence such explosive RI (ERI). Using the coupled Weather Research and Forecasting and 3D Price–Weller–Pinkel Ocean model, we compared a Hagibis control run with two sensitivity experiments using non-MHW sea surface temperatures (SSTs) and vertical temperature profiles from another RI case of Typhoon Jebi (2018). Hagibis intensifies dramatically faster under MHW conditions—about 20% faster in the baseline simulation, and 40% faster when realistic vertical ocean mixing in the surface layer is included—revealing the striking sensitivity of ERI to upper-ocean thermal anomalies. Strong latent heat fluxes supplied continuous energy to the lower troposphere, while elevated subsurface heat content sustained air-sea enthalpy exchange despite storm-induced SST cooling. These ocean-atmosphere interactions reorganized TC convection, strengthened updrafts, and fostered an environment conducive to ERI. We conclude that even a 1 °C increase in surface temperature is sufficient to modulate storm structure and intensity, amplifying ERI. As ocean temperatures continue to rise, these findings highlight the need to refine the definition of ERI and improve our understanding of TC-ocean interactions under increasing MHW influence.
- 30 April 2026
- Published by IOP Publishing in Environmental Research Letters
- Vol. 21 (10) , 104013
Open AccessModest ocean warming difference, explosive intensification: Marine heatwave control of Typhoon Hagibis (2019)
Typhoon Hagibis (2019) explosively intensified by 100 kt in 24 hours, greatly exceeding the conventional rapid intensification (RI) threshold of ~30 kt per 24 h. This study investigates how marine heatwave (MHW)–driven ocean surface and subsurface structures influence such explosive RI (ERI). Using the coupled Weather Research and Forecasting and 3D Price–Weller–Pinkel Ocean model, we compared a Hagibis control run with two sensitivity experiments using non-MHW sea surface temperatures (SSTs) and vertical temperature profiles from another RI case of Typhoon Jebi (2018). Hagibis intensifies dramatically faster under MHW conditions—about 20% faster in the baseline simulation, and 40% faster when realistic vertical ocean mixing in the surface layer is included—revealing the striking sensitivity of ERI to upper-ocean thermal anomalies. Strong latent heat fluxes supplied continuous energy to the lower troposphere, while elevated subsurface heat content sustained air-sea enthalpy exchange despite storm-induced SST cooling. These ocean-atmosphere interactions reorganized TC convection, strengthened updrafts, and fostered an environment conducive to ERI. We conclude that even a 1 °C increase in surface temperature is sufficient to modulate storm structure and intensity, amplifying ERI. As ocean temperatures continue to rise, these findings highlight the need to refine the definition of ERI and improve our understanding of TC-ocean interactions under increasing MHW influence.
- Top Cited PapersOpen Access
- 7 February 2024
- Published by Springer Nature in Communications Earth & Environment
- Vol. 5 (1) , 1-8
Marine heatwave events strengthen the intensity of tropical cyclones
Marine heatwaves become more frequent and stronger due to global warming. It is necessary to clarify how and by which process marine heatwaves affect marine weather systems. Here, we examine how marine heatwaves strengthen tropical cyclones. We analyze 128 tropical cyclones that intensified with marine heatwaves and 184 tropical cyclones that intensified without marine heatwaves over the western North Pacific and Atlantic. Marine heatwaves finally lead to a maximum intensity of 35.4% stronger (106.72 kts) tropical cyclone. Notably, satellite data reveal that marine heatwaves are associated with precipitation-richer tropical cyclone conditions despite similar tropical cyclone intensities. Due to the increased latent heat flux during marine heatwaves, intensified precipitation near the tropical cyclone center contributes to the intensification process. This study provides insights into how more frequent marine heatwaves by global warming impact tropical cyclones.
- 7 February 2024
- Published by Springer Nature in Communications Earth & Environment
- Vol. 5 (1) , 1-8
Top Cited PapersOpen AccessMarine heatwave events strengthen the intensity of tropical cyclones
Marine heatwaves become more frequent and stronger due to global warming. It is necessary to clarify how and by which process marine heatwaves affect marine weather systems. Here, we examine how marine heatwaves strengthen tropical cyclones. We analyze 128 tropical cyclones that intensified with marine heatwaves and 184 tropical cyclones that intensified without marine heatwaves over the western North Pacific and Atlantic. Marine heatwaves finally lead to a maximum intensity of 35.4% stronger (106.72 kts) tropical cyclone. Notably, satellite data reveal that marine heatwaves are associated with precipitation-richer tropical cyclone conditions despite similar tropical cyclone intensities. Due to the increased latent heat flux during marine heatwaves, intensified precipitation near the tropical cyclone center contributes to the intensification process. This study provides insights into how more frequent marine heatwaves by global warming impact tropical cyclones.
- Open Access
- 20 January 2024
- Published by Springer Nature in npj Climate and Atmospheric Science
- Vol. 7 (1) , 1-7
Sea surface temperature warming to inhibit mitigation of tropical cyclone destructiveness over East Asia in El Niño
Given their conditions to reside in and intensify longer over warm oceans, tropical cyclones (TCs) in the western North Pacific (WNP) present a stronger lifetime maximum intensity during El Niño than during La Niña. By using observational data, we found that the anomalously cool sea surface temperature (SST)s in the basin act as effective barriers against intense TCs approaching East Asia during El Niño, weakening their destructiveness at landfall. Based on our high-resolution pseudo-global-warming simulations, the basin-wide 2K SST warming within the WNP basin can, however, shatter this cool SST barrier, exposing East Asia to more destructive TCs during El Niño, compared to those during La Niña. Considering that the 2K warmer WNP will likely occur in the mid-21st century under a high emission scenario and in the late 21st century under a moderate emission scenario, our findings support that more aggressive efforts of global warming mitigation are needed.
- 20 January 2024
- Published by Springer Nature in npj Climate and Atmospheric Science
- Vol. 7 (1) , 1-7
Open AccessSea surface temperature warming to inhibit mitigation of tropical cyclone destructiveness over East Asia in El Niño
Given their conditions to reside in and intensify longer over warm oceans, tropical cyclones (TCs) in the western North Pacific (WNP) present a stronger lifetime maximum intensity during El Niño than during La Niña. By using observational data, we found that the anomalously cool sea surface temperature (SST)s in the basin act as effective barriers against intense TCs approaching East Asia during El Niño, weakening their destructiveness at landfall. Based on our high-resolution pseudo-global-warming simulations, the basin-wide 2K SST warming within the WNP basin can, however, shatter this cool SST barrier, exposing East Asia to more destructive TCs during El Niño, compared to those during La Niña. Considering that the 2K warmer WNP will likely occur in the mid-21st century under a high emission scenario and in the late 21st century under a moderate emission scenario, our findings support that more aggressive efforts of global warming mitigation are needed.
- Open Access
- 16 March 2022
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 49 (5)
Latitudinal Variation of the Lifetime Maximum Intensity Location of Atlantic Tropical Cyclones Controlled by the Atlantic Multidecadal Oscillation
The latitude of lifetime maximum intensity (LMI) location (ϕLMI) of Atlantic tropical cyclones (TCs) is related to the Atlantic Multidecadal Oscillation (AMO) with a significant negative correlation for 1948–2018. We objectively classify TC tracks into four patterns, to evaluate the effects of variations in dominant track patterns (track ϕLMI) and intra‐pattern ϕLMI (pure ϕLMI) on the total ϕLMI variation. During the warm phases of AMO, the prevailed south‐origin track patterns and equatorward pure ϕLMI shifts have almost equally contributed to the equatorward migration of the total ϕLMI. The negative anomaly of potential intensity in the subtropics is responsible for the equatorward shift of pure ϕLMI. Moreover, weak vertical wind shear in the tropics increases the south‐origin track patterns. These environmental conditions are associated with the east‐strong and west‐weak relative sea surface temperature warming during the warm phases that causes anomalous upward and downward motions in the eastern and western Atlantic, respectively.
- 16 March 2022
- Published by American Geophysical Union (AGU) in Geophysical Research Letters
- Vol. 49 (5)
Open AccessLatitudinal Variation of the Lifetime Maximum Intensity Location of Atlantic Tropical Cyclones Controlled by the Atlantic Multidecadal Oscillation
The latitude of lifetime maximum intensity (LMI) location (ϕLMI) of Atlantic tropical cyclones (TCs) is related to the Atlantic Multidecadal Oscillation (AMO) with a significant negative correlation for 1948–2018. We objectively classify TC tracks into four patterns, to evaluate the effects of variations in dominant track patterns (track ϕLMI) and intra‐pattern ϕLMI (pure ϕLMI) on the total ϕLMI variation. During the warm phases of AMO, the prevailed south‐origin track patterns and equatorward pure ϕLMI shifts have almost equally contributed to the equatorward migration of the total ϕLMI. The negative anomaly of potential intensity in the subtropics is responsible for the equatorward shift of pure ϕLMI. Moreover, weak vertical wind shear in the tropics increases the south‐origin track patterns. These environmental conditions are associated with the east‐strong and west‐weak relative sea surface temperature warming during the warm phases that causes anomalous upward and downward motions in the eastern and western Atlantic, respectively.
- Open Access
- 17 September 2021
- Published by MDPI AG in Atmosphere
- Vol. 12 (9) , 1214
A Performance Evaluation of Potential Intensity over the Tropical Cyclone Passage to South Korea Simulated by CMIP5 and CMIP6 Models
Potential intensity (PI) is a metric for climate model evaluation of TC-related thermodynamic conditions. However, PI is utilized usually for assessing basin-wide TC-related thermodynamic conditions, and not for evaluating TC passage to a certain region. Here we evaluate model-simulated PI over the passage of TCs affecting South Korea (KOR PI) as well as the PI over the entire western North Pacific basin (WNP PI) using 25 CMIP5 and 27 CMIP6 models. In terms of pattern correlations and bias-removed root mean square errors, CMIP6 model performances for KOR PI are found to be noticeably improved over CMIP5 models in contrast to negligible improvement for WNP PI, although it is not in terms of normalized standard deviations. This implies that thermodynamic condition on the route of TCs affecting South Korea is likely better captured by CMIP6 models than CMIP5 models.
- 17 September 2021
- Published by MDPI AG in Atmosphere
- Vol. 12 (9) , 1214
Open AccessA Performance Evaluation of Potential Intensity over the Tropical Cyclone Passage to South Korea Simulated by CMIP5 and CMIP6 Models
Potential intensity (PI) is a metric for climate model evaluation of TC-related thermodynamic conditions. However, PI is utilized usually for assessing basin-wide TC-related thermodynamic conditions, and not for evaluating TC passage to a certain region. Here we evaluate model-simulated PI over the passage of TCs affecting South Korea (KOR PI) as well as the PI over the entire western North Pacific basin (WNP PI) using 25 CMIP5 and 27 CMIP6 models. In terms of pattern correlations and bias-removed root mean square errors, CMIP6 model performances for KOR PI are found to be noticeably improved over CMIP5 models in contrast to negligible improvement for WNP PI, although it is not in terms of normalized standard deviations. This implies that thermodynamic condition on the route of TCs affecting South Korea is likely better captured by CMIP6 models than CMIP5 models.
- Open Access
- 18 August 2021
- Published by MDPI AG in Atmosphere
- Vol. 12 (8) , 1061
Effect of Mid-Latitude Jet Stream on the Intensity of Tropical Cyclones Affecting Korea: Observational Analysis and Implication from the Numerical Model Experiments of Typhoon Chaba (2016)
The effect of the jet stream on the changes in the intensity of tropical cyclones (TC) affecting Korea is discussed. We classified the TCs into three categories based on the decreasing rate of TC intensity in 24 h after TC passed 30° N. The TCs with a large intensity decrease had a more vigorous intensity when the TCs approached the mid-latitudes. The analysis of observational fields showed that the strong jet stream over Korea and Japan may intensify TCs by the secondary circulations of jet entrance but induces a large decrease in TC intensity in the mid-latitudes by the strong vertical wind shear. We also performed the numerical simulation for the effect of the jet stream on the intensity changes of Typhoon Chaba (2016). As a result, the stronger jet stream induced more low-level moisture convergence at the south of the jet stream entrance, enhancing the intensity when the TC approached Korea. Furthermore, it induced a rapid reduction in intensity when TC approached in the strong jet stream area. The results suggest that the upper-level jet stream is one of the critical factors modulating the intensity of TC affecting Korea in the vicinity of the mid-latitudes.
- 18 August 2021
- Published by MDPI AG in Atmosphere
- Vol. 12 (8) , 1061
Open AccessEffect of Mid-Latitude Jet Stream on the Intensity of Tropical Cyclones Affecting Korea: Observational Analysis and Implication from the Numerical Model Experiments of Typhoon Chaba (2016)
The effect of the jet stream on the changes in the intensity of tropical cyclones (TC) affecting Korea is discussed. We classified the TCs into three categories based on the decreasing rate of TC intensity in 24 h after TC passed 30° N. The TCs with a large intensity decrease had a more vigorous intensity when the TCs approached the mid-latitudes. The analysis of observational fields showed that the strong jet stream over Korea and Japan may intensify TCs by the secondary circulations of jet entrance but induces a large decrease in TC intensity in the mid-latitudes by the strong vertical wind shear. We also performed the numerical simulation for the effect of the jet stream on the intensity changes of Typhoon Chaba (2016). As a result, the stronger jet stream induced more low-level moisture convergence at the south of the jet stream entrance, enhancing the intensity when the TC approached Korea. Furthermore, it induced a rapid reduction in intensity when TC approached in the strong jet stream area. The results suggest that the upper-level jet stream is one of the critical factors modulating the intensity of TC affecting Korea in the vicinity of the mid-latitudes.
- 1 May 2017
- Published by American Meteorological Society in Journal of Climate
- Vol. 30 (9) , 3297-3302
Variation of the Tropical Cyclone Season Start in the Western North Pacific
The variation of the tropical cyclone (TC) season in the western North Pacific (WNP) was analyzed based on the percentiles of annual TC formation dates. The results show that the length of the TC season is highly modulated by the TC season’s start rather than its end. The start of the TC season in the WNP has large interannual variation that is closely associated with the variation of the sea surface temperature (SST) in the Indian Ocean (IO) and the central-eastern Pacific (CEP). When the SSTs of the IO and CEP are warm (cold) in the preceding winter, anomalous high (low) pressure and anticyclonic (cyclonic) circulation are induced around the WNP TC basin the following spring, resulting in a late (early) start of the TC season. These results suggest that a strong El Niño in the preceding winter significantly delays the TC season start in the following year.
- 1 May 2017
- Published by American Meteorological Society in Journal of Climate
- Vol. 30 (9) , 3297-3302
Variation of the Tropical Cyclone Season Start in the Western North Pacific
The variation of the tropical cyclone (TC) season in the western North Pacific (WNP) was analyzed based on the percentiles of annual TC formation dates. The results show that the length of the TC season is highly modulated by the TC season’s start rather than its end. The start of the TC season in the WNP has large interannual variation that is closely associated with the variation of the sea surface temperature (SST) in the Indian Ocean (IO) and the central-eastern Pacific (CEP). When the SSTs of the IO and CEP are warm (cold) in the preceding winter, anomalous high (low) pressure and anticyclonic (cyclonic) circulation are induced around the WNP TC basin the following spring, resulting in a late (early) start of the TC season. These results suggest that a strong El Niño in the preceding winter significantly delays the TC season start in the following year.