- Research Letter
- Open Access
Change in the tropical cyclone activity around Korea by the East Asian summer monsoon
© The Author(s) 2017
- Received: 2 September 2016
- Accepted: 6 January 2017
- Published: 21 January 2017
Correlation between the frequency of summer tropical cyclones (TCs) affecting Korea and the East Asian summer monsoon index (EASMI) was analyzed over the last 37 years. A clear positive correlation existed between the two variables, and this high positive correlation remained unchanged even when excluding El Niño-Southern Oscillation (ENSO) years. To investigate the causes of the positive correlation between the two variables in non-ENSO years, after the 8 years with the highest EASMI (high EASMI years) and the 8 years with the lowest EASMI (low EASMI years) were selected, and the average difference between the two phases was analyzed. In high EASMI years, in the difference between the two phases regarding 850 and 500 hPa streamline, anomalous cyclones were reinforced in the tropical and subtropical western North Pacific, while anomalous anticyclones were reinforced in mid-latitude East Asian areas. Due to these two anomalous pressure systems, anomalous southeasterlies developed near Korea, with these anomalous southeasterlies playing the role of anomalous steering flows making the TCs head toward areas near Korea. In addition, a monsoon trough strengthened more eastward, and TCs in high EASMI years occurred more in east ward over the western North Pacific.
- Tropical cyclones
- East Asian summer monsoon
While many studies performed on changes in TCs in individual tropical cyclone (TC) genesis basin or major regions affected by TC activity, studies conducted on TCs that affected or made landfall on Korea are rare. Choi and Kim (2007) showed that the frequency of TCs that made landfall on Korea had been increasing since the middle of the 1990s and increases in the landfalling frequency of strong TCs had been more prominent. As for these results, they analyzed that the reason was that as the western North Pacific subtropical high (WNPSH) shrunk toward the east, the paths of TCs moved to the east so that the frequency of passing through China before making landfall on Korea decreased. Choi et al. (2010) applied the statistical change-point analysis used in a study by Ho et al. (2004) to changes in the frequency of TCs that made landfall on Korea. Based on the results, they showed that the analysis period for the past 54 years (1951–2004) could be divided into three periods and that the landfalling frequency of strong TCs was the highest in the most recent period. Park et al. (2006) showed that this increasing tendency was clear in TCs that affected Korea. In particular, Choi and Kim (2011) showed that the frequency of TCs that affected Korea increased since the early 1980s after applying the statistical change-point analysis.
However, the frequency of TCs that affected Korea decreased abnormally since the late 2000s so that TC season forecasters have been experiencing considerable difficulties when they were forecasting TCs. Furthermore, since the above studies used dataset obtained only until the early 2000, opinions indicating that the activities of TCs that affected or made landfall on Korea should be investigated again using dataset with more long period have been raised (Choi and Kim 2011; Choi et al. 2009).
The East Asia region has been suffering from enormous property and human life damage every year due to TCs. In Korea, after the landing of typhoon ‘RUSA’ in 2002, the highest record (the 24 h accumulated precipitation of 870.5 mm) since the time when weather observation began in Korea was observed in Gangneung (Park and Lee 2007). In addition to Korea, the highest records of strong winds, heavy rains, and waves that accompanied by TCs have been renewed in Japan, Taiwan, and Philippines (Kim et al. 2005; Lyon and Camargo 2009; Pan et al. 2010).
Efforts have been actively made to find the cause of these abnormal TC activities occurring in various East Asia regions using climate factors. Among those, climate factor that has received the most keen attention thus far in relation to efforts to understand the characteristics of TC activity is El Niño–Southern Oscillation (ENSO) (Wang and Chan 2002; Chu 2004; Wu et al. 2004). In general, TCs in El Niño phase mainly occur in the southeast quadrant of western North Pacific and tend to make landfall on or affect the east coast of China, Korea, or Japan. On the other hand, TCs in La Niña phase occur in the northwest quadrant of western North Pacific and tend to move westward or northwestward. Therefore, TCs mainly make landfall on Philippines, southern China, or Indochina Peninsula. Therefore, TCs in El Niño phase show a little longer life-span and a little higher intensity than those in La Niño phase. Meanwhile, Ho et al. (2005) tried to associate the cause of changes in TC activity in the western North Pacific with atmospheric circulations in the Southern Hemisphere. They showed that during the positive Antarctic Oscillation (AAO) phase, TC passage frequency increased due to WNPSH reinforced in the East Asian middle latitude region but decreased in regions in the vicinity of the South China Sea. Wang and Fan (2007) found that AAO and TC genesis frequency in the western North Pacific had negative correlations during June to September (JJAS). In that study, they showed that during the positive AAO phase, vertical wind shear increased, the vertical structure of lower anticyclone/upper cyclone developed in the western North Pacific, and sea surface temperatures dropped so that unfavorable environments for the occurrence of TCs were formed. Meanwhile, Wang et al. (2007) investigated the relationship between the North Pacific Oscillation (NPO) and TC genesis frequency in the western North Pacific and the subtropical Atlantic during JJAS. Based on the results, they suggested that TC genesis frequency and the NPO showed positive correlations and negative correlations in the former and latter basins, respectively, and that changes in TC genesis frequency between the two basins were made through teleconnection patterns. As the recent study, Choi and Byun (2010) analyzed changes in TC activity in the western North Pacific according to Pacific-Japan teleconnection patterns (PJ pattern). They showed that during positive PJ patterns, because anomalous cyclonic circulation and anomalous anticyclonic circulation were reinforced in low and middle latitude regions of East Asian, respectively, the anomalous southerlies reinforced in the mid-latitude regions played the role of steering flows to have TCs move to these regions more easily. However, studies on the relationship between the Arctic Oscillation (AO) which is one of the most clear zonal circulation (or annular) modes in winter in the Northern Hemisphere and TC activity in the western North Pacific cannot be easily found. Only Choi and Byun (2010) analyzed changes in TC activity in the western North Pacific according to AO phases during July to September. Studies on the relationship between TC activity and the AO in the Atlantic are also rare (Larson et al. 2005; Xie et al. 2005).
However, most of studies on TCs that affected or made landfall on Korea have been concentrated on the precipitation and strong winds that accompanied by TCs or their modeling due to the property and human life damage (Kim and Lee 2007; Kim et al. 2010; Park and Kim 2010). However, understanding the trend of the annual frequency of TCs that affected or made landfall on Korea and those climate factors that cause the trend should be also helpful for seasonal forecasting of TCs or preparation for disasters resulting from TCs. Therefore, the present study examines the effects of East Asian summer monsoon on changes in the frequency of TCs that affect Korea.
In the present study, materials and analysis methods are introduced in “Data and methods” section, and East Asian summer monsoons are defined in “Relationship between TC frequency around Korea and EASM” section. In “Differences between high EASMI years and low EASMI years” section, the relationships between the frequency of TCs that affect Korea and East Asian summer monsoon are analyzed, and in “Summary and conclusion” section, large-scale environments are examined to figure out the cause of these relationships. The present study is summarized in “Relationship between TC frequency around Korea and EASM” section.
The TC datasets in this study were obtained from the TC best track provided by Regional Specialized Meteorological Center (RSMC)-Tokyo Typhoon Center. These datasets consist of TC name, latitude, and longitude location of TC, TC central pressure, and TC Maximum Sustained Wind Speed (MSWS), which were observed in every 6 h for 37 years (1977–2013). TC is generally classified into four classes by the criteria of MSWS as follows: Tropical Depression (TD, MSWS < 17 m s−1), Tropical Storm (TS, 17 m s−1 ≤ MSWS ≤ 24 m s−1), Severe Tropical Storm (STS, 25 m s−1 ≤ MSWS ≤ 32 m s−1), and Typhoon (TY, MSWS ≥ 33 m s−1). Along with the four classes of TC above, this study included extratropical cyclone, which was transformed from TC. This was because such extratropical cyclone also incurred great damage on property and human in the mid-latitude regions of East Asia.
Moreover, this study also used the variables of geopotential height (gpm), zonal and meridional winds (m s−1), precipitable water (kgm−2), and relative humidity (%) from National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis in 1977–2013 (Kalnay et al. 1996; Kistler et al. 2001). This NCEP-NCAR reanalysis dataset consisted of spatial resolution such as latitude and longitude 2.5° × 2.5° and 17 vertical levels (relative humidity is 16 vertical levels and precipitable water is 1 level). Also, velocity potential consisted of a spatial resolution such as latitude and longitude 192 × 94 and 5 sigma levels.
The NOAA Extended Reconstructured monthly Sea Surface Temperature (SST) (Reynolds et al. 2002), available from the same organization, was also used. The datasets have a horizontal resolution of 2.0° × 2.0° latitude–longitude and are available for the period of 1854 to the present day.
In addition, the Climate Prediction Center (CPC) Merged Analysis of Precipitation (CMAP) dataset (Xie and Arkin 1997), the horizontal spatial resolution of which is the same as the NCEP–NCAR reanalysis dataset. These datasets are based on the monthly average and are available from 1979 to the present day. The CMAP dataset, which is global precipitation data that covers the ocean, are derived by merging rain gauge observations, five different satellite estimates, and numerical model outputs.
In order to calculate TC passage frequencies, each TC was calculated after being relocated within a 5° × 5° grid. Even if a TC passed over the same grid multiple times, it was regarded as a single passage. TC genesis frequencies were also calculated in the above manner.
TC intensity is defined as the TC central pressure and TC MSWS observed in the vicinity of Korea using the TC best track data at intervals of 6 h provided by the RSMC-Tokyo typhoon center.
Here, u and v indicate the zonal and meridional flows, respectively. 200 and 850 represent 200 and 850 hPa levels, respectively (Wingo and Cecil 2010).
In this study, summer is defined as the period from June to August. This is because EASM rainfall is concentrated on these 3 months (Wang and Fan 1999).
Definition of EASM index (EASMI)
Results of partial correlation analysis
TC frequency around Korea on high EASMI years and low EASMI years
High EASMI years
Low EASMI years
TC genesis frequency on high EASMI years and low EASMI years
High EASMI years
Low EASMI years
In addition, differences in TC passage frequencies between the two groups were analyzed (Fig. 4b). In high EASMI years, TCs showed a tendency to move from the far southeastern sea of the Philippines, pass the East China Sea, and move northward toward Korea and Japan. On the other hand, in low EASMI years, TCs showed a tendency to move from the southeastern sea of the Philippines, pass the South China Sea, and make landfall on the Indochina Peninsula. Earlier, it was identified that TCs in high EASMI years mainly occurred in the east waters of the western North Pacific. Wang and Chan (2002) showed that TCs that occurred in the east waters of the western North Pacific showed a tendency to move northward toward mid-latitudes of East Asia, while TCs that occurred in the west waters had a strong tendency to move westward toward the southeastern region of China and the Indochina Peninsula. The results in the present study were consistent with the results of Wang and Chan (2002). Eventually, it can be seen that dipole patterns appear between the southeastern region and the northeastern region of East Asia from the difference in TC passage frequencies between the two groups.
To examine the cause of the positive correlation between EASM and TC frequency around Korea identified earlier, differences in large-scale environments between the two groups were analyzed.
The degree of development of monsoon troughs in the two groups is associated with the degree of development of WNPSHs. Therefore, the degrees of development of WNPSHs in the two groups were analyzed (Fig. 4b). Here, WNPSHs are defined as areas having values larger than 5875 gpm at 500 hPa. The WNPSH in low EASMI years when the monsoon trough was weak expanded westward to the southeastern coast of China. On the other hand, the WNPSH in high EASMI years when the monsoon trough was strong were weakened eastward. In general, TCs tend to move along the western periphery of WNPSHs (Wang and Chan 2002). Therefore, the TC track of each group was almost identical to the western periphery of WNPSH, and TCs moved a little further to regions around Korea in high EASMI years when the WNPSH shrunk eastward.
On the other hand, Fig. 4a shows that TC formation increases in the 130°–150°E, 5°–20°N region during the high EASMI years. They move with a recurving track along the western boundary of an eastward retreated Pacific subtropical high toward Korea. In the low EASMI years, TC formation tends to increase in the oceans to the east/southeast of Taiwan and affected by a westward expanded WNPSH to move westward toward the South China Sea. These analyses reveal two important features: variability of TC formation location and moving tracks. The composite 850 hPa streamline in Fig. 8 explains the different steering effects of the WNPSH on the result in two different TC tracks between the high and low EASMI years. Also, composite anomalies of horizontal divergence (Fig. 9), relative humidity, and vertical wind shear (Fig. 12) explain the variability of TC formation.
With regard to differences in 200–850 hPa vertical wind shear between the two groups, negative anomalies were shown in all waters except for west waters of the western North Pacific and in the mid-latitudes of East Asia in high EASMI years (Fig. 12b). When the value of 200–850 hPa vertical wind shear is smaller, differences in wind directions between the upper and lower layers of the troposphere are smaller so that TCs can occur easily, and high TC intensity can be maintained. Therefore, in high EASMI years, more TCs can occur in the western North Pacific, and high TC intensity can be maintained even when TCs have moved northward to regions around Korea.
In the present study, the correlations between the frequency of TCs that affected Korea in summer over the last 37 years and the EASMI defined by Li and Zeng (2002, 2003, 2005) were analyzed. Clear positive correlations existed between the two variables and the high positive correlations did not change even when ENSO years had been excluded. To examine the cause of these positive correlations between the two variables, 8 years having the highest EASMI (high EASMI years) and 8 years having the lowest EASMI (low EASMI years) except for ENSO years were selected, and differences in averages between the two groups were analyzed.
In high EASMI years, TCs mainly occurred in the east waters of the western North Pacific and showed a tendency to pass the East China Sea and move northward toward regions around Korea. In low EASMI years, TCs mainly occurred in the west sea of the western North Pacific and showed a pattern of passing the South China Sea and moving westward toward the Indochina Peninsula. Therefore, TC intensity was higher in high EASMI years when TCs could obtain sufficient energy while moving long distances to regions around Korea. In addition, TCs showed a characteristic of occurring more frequently in high EASMI years than in low EASMI years.
With regard to differences in 850 and 500 hPa streamlines between the two groups, anomalous cyclones were reinforced in the western North Pacific and anomalous anticyclones were reinforced in the mid-latitudes of East Asia in high EASMI years. Due to these two anomalous pressure systems, anomalous southeasterlies developed around Korea, and these anomalous southeasterlies played the role of anomalous steering flows to have TCs move toward regions around Korea. In addition, more TCs could occur in high EASMI years because of the anomalous cyclones developed in the western North Pacific.
The anomalous cyclones developed in the western North Pacific in high EASMI years were associated with the development of monsoon troughs. While the monsoon trough developed eastward to 150°E in high EASMI years, they hardly developed in low EASMI years. Therefore, some more TCs could occur in high EASMI years mainly in the east waters of the western North Pacific.
Differences in the degree of development of monsoon troughs between the two groups were associated with the degree of development of WNPSHs. While the WNPSH shrunk eastward in high EASMI years, it expanded to the southeastern coast of China in low EASMI years. Therefore, while TCs could move northward to regions around Korea in high EASMI years, they became to move westward toward the Indochina Peninsula in low EASMI years.
Meanwhile, anomalous anticyclones and anomalous cyclones developed in the mid-latitudes of East Asia and the western North Pacific, respectively, at the lower layer of the troposphere in high EASMI years could be identified through differences in horizontal divergences between the two groups. At the lower layer of the troposphere, anomalous convergences were reinforced in the western North Pacific and anomalous divergences were reinforced in the mid-latitudes of East Asia. At the upper layer of the troposphere, the opposite pattern was reinforced so that anomalous upward flows were formed in the western North Pacific, while anomalous downward flows were formed in the mid-latitudes of East Asia.
To examine whether these anomalous vertical atmospheric circulations were reinforced in high EASMI years, those circulations in 100°–180°E where TCs mainly occur were averaged to analyze differences in vertical meridional atmospheric circulations between the two groups. Anomalous secondary circulations were formed in which air currents ascended in regions south to 25°N and descended in 25°–40°N. Because of the anomalous vertical meridional atmospheric circulations reinforced as such in high EASMI years, positive relative humidity anomalies were reinforced at all layers of the troposphere in regions south to 30°N, while negative relative humidity anomalies were reinforced at all layers of the troposphere in 30°–45°N. These conditions of anomalous vertical meridional atmospheric circulations and relative humidity became favorable environments for more TCs to occur in high EASMI years and have more effects on regions around Korea.
Both the upper and lower layers of the troposphere showed warm anomalies in most regions in the analysis area. In the middle layer relative humidity of the troposphere, positive anomalies were shown in the western North Pacific to provide favorable environments for TC intensity to be reinforced further in high EASMI years. In addition, in 200–850 hPa vertical wind shear, not only the western North Pacific but also the mid-latitudes of East Asia showed negative anomalies and mid-latitude waters of East Asia showed warm SST anomalies so that favorable environments were formed for TC intensity to be reinforced further in high EASMI years.
According to analyses of global-scale atmospheric circulations, in high EASMI years, the converged at the lower layer of the western North Pacific was diverged at the upper layer and the diverged air moved westward to converge at the upper layer of the equatorial Indian Ocean and diverge at the lower layer.
In the present study, the effects of EASM on TC activity around Korea were analyzed. These results are considered helpful for TC seasonal prediction for regions around Korea. In future studies, statistical TC seasonal prediction models will be developed using EASM.
JWC designed and carried out the study, and wrote the paper. YC provided detailed information on East Asian meteorological data. JYK acquired and processed the data of YC. JWC, YC, and JYK contributed extensively to the scientific discussion. All authors read and approved the final manuscript.
The authors declare that they have no competing interests.
This work was supported by the R&D Project of the Korea Meteorological Administration “Development and application of technology for weather forecast.”
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
- Choi KS, Byun HR (2010) Possible relationship between western North Pacific tropical cyclone activity and Arctic Oscillation. Theor Appl Climatol 100:261–274View ArticleGoogle Scholar
- Choi KS, Kim BJ (2007) Climatological characteristics of tropical cyclones making landfall over the Korean Peninsula. Asia Pacific J Atmos Sci 43:97–109Google Scholar
- Choi KS, Kim TR (2011) Regime shift of the early 1980s in the characteristics of the tropical cyclone affecting Korea. J Korean Earth Sci Soc 32:453–460View ArticleGoogle Scholar
- Choi KS, Kim DW, Byun HR (2009) Statistical model for seasonal prediction of tropical cyclone frequency around Korea. Asia Pacific J Atmos Sci 45:21–32Google Scholar
- Choi KS, Kim BJ, Kim DW, Byun HR (2010) Interdecadal variation of tropical cyclone making landfall over the Korean Peninsula. Int J Climatol 30:1472–1483Google Scholar
- Chu PS (2004) ENSO and tropical cyclone activity. In: Murnane RJ, Liu K-B (eds) Hurricanes and typhoons: past, present, and potential. Columbia University Press, New York, pp 297–332Google Scholar
- Gray WM (1975) Tropical cyclone genesis. Dept. of Atmospheric Science Paper 234, Colorado State University, Fort CollinsGoogle Scholar
- Ho CH, Baik JJ, Kim JH, Gong DY, Sui CH (2004) Interdecadal changes in summertime typhoon tracks. J Clim 17:1767–1776View ArticleGoogle Scholar
- Ho CH, Kim JH, Kim HS, Sui CH, Gong DY (2005) Possible influence of the Antarctic Oscillation on tropical cyclone activity in the western North Pacific. J Geophys Res 110:D19104. doi:10.1029/2005JD005766 View ArticleGoogle Scholar
- Kalnay E, Kanamitsu M, Kistler R et al (1996) The NCEP/NCAR 40-year reanalysis project. Bull Am Meteorol Soc 77:437–471View ArticleGoogle Scholar
- Kim JW, Lee JG (2007) A qualitative analysis of WRF simulation results of typhoon `Rusa’ Case. Atmosphere 17:393–405Google Scholar
- Kim JH, Ho CH, Sui CH (2005) Circulation features associated with the record-breaking typhoon landfall on Japan in 2004. Geophy Res Lett 32:L14713. doi:10.1029/2005GL022494 View ArticleGoogle Scholar
- Kim YH, Jeon EH, Chang DE, Lee HS, Park JI (2010) The impact of T-PARK 2008 dropsonde observations on typhoon track forecasting. Asia Pacific J Atmos Sci 46:287–303View ArticleGoogle Scholar
- Kistler R, Kalnay E, Collins W et al (2001) The NCEP–NCAR 50-year reanalysis: monthly means CD-ROM and documentation. Bull Am Meteor Soc 82:247–267View ArticleGoogle Scholar
- Larson J, Zhou Y, Higgins RW (2005) Characteristics of landfalling tropical cyclones in the United States and Mexico: climatology and interannual variability. J Clim 18:1247–1262View ArticleGoogle Scholar
- Li JP, Zeng QC (2002) A unified monsoon index. Geophy Res Lett 29:1274. doi:10.1029/2001GL013874 Google Scholar
- Li JP, Zeng QC (2003) A new monsoon index and the geographical distribution of the global monsoons. Adv Atmos Sci 20:299–302View ArticleGoogle Scholar
- Li JP, Zeng QC (2005) A new monsoon index, its interannual variability and relation with monsoon precipitation. Clim Environ Res 10:351–365Google Scholar
- Lyon B, Camargo SJ (2009) The seasonally-varying influence of ENSO on rainfall and tropical cyclone activity in the Philippines. Clim Dyn 32:125–141View ArticleGoogle Scholar
- Nitta T (1986) Long-term variations of cloud amount in the western Pacific region. J Meteor Soc Jpn 64:373–390Google Scholar
- Nitta T (1987) Convective activities in the tropical western Pacific and their impact on the Northern Hemisphere summer circulation. J Meteor Soc Japan 65:373–390Google Scholar
- Nitta T (1989) Global features of the Pacific-Japan oscillation. Meteor Atmos Phys 41:5–12View ArticleGoogle Scholar
- Pan CJ, Reddy KK, Lai HC, Yang SS (2010) Role of mixed precipitating cloud systems on the typhoon rainfall. Ann Geophys 28:11–16View ArticleGoogle Scholar
- Park JI, Kim HM (2010) Typhoon Wukong (200610) Prediction based on the ensemble Kalman filter and ensemble sensitivity analysis. Atmosphere 20:287–306Google Scholar
- Park SK, Lee EH (2007) Synoptic features of orographically enhanced heavy rainfall on the east coast of Korea associated with Typhoon Rusa (2002). Geophys Res Lett 34:L02803. doi:10.1029/2006GL028592 Google Scholar
- Park JK, Km BS, Jung WS, Kim EB, Lee DG (2006) Change in statistical characteristics of typhoon affecting the Korean Peninsula. Atmosphere 16:1–17Google Scholar
- Ramage CS (1971) Monsoon meteorology. Academic Press, New YorkGoogle Scholar
- Reynolds RW, Rayner NA, Smith TM, Stokes DC, Wang W (2002) An improved in situ and satellite SST analysis for climate. J Clim 15:1609–1625View ArticleGoogle Scholar
- Wang B, Chan JCL (2002) How strong ENSO events affect tropical storm activity over the western North Pacific. J Clim 15:1643–1658View ArticleGoogle Scholar
- Wang B, Fan Z (1999) Choice of South Asian summer monsoon indices. Bull Am Meteor Soc 80:629–638View ArticleGoogle Scholar
- Wang HJ, Fan K (2007) Relationship between the Antarctic Oscillation in the western North Pacific typhoon frequency. Chin Sci Bull 52:561–565View ArticleGoogle Scholar
- Wang HJ, Sun JQ, Ke F (2007) Relationships between the north Pacific oscillation and the typhoon/hurricane frequencies. Sci China, Ser D Earth Sci 50:1409–1416View ArticleGoogle Scholar
- Wilks DS (1995) Statistical methods in the Atmospheric Sciences. Academic Press, New YorkGoogle Scholar
- Wingo MT, Cecil DJ (2010) Effects of vertical wind shear on tropical cyclone precipitation. Mon Wea Rev 138:645–662View ArticleGoogle Scholar
- Wu MC, Chang WL, Leung WM (2004) Impacts of El Niño-Southern Oscillation events on tropical cyclone landfalling activity in the western North Pacific. J Clim 15:1419–1428View ArticleGoogle Scholar
- Xie P, Arkin PA (1997) Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs. Bull Am Meteorol Soc 78:2539–2558View ArticleGoogle Scholar
- Xie L, Yan T, Pietrafesa LJ (2005) Climatology and interannual variability of North Atlantic hurricane tracks. J Clim 18:5370–5381View ArticleGoogle Scholar