Climatological and Seasonal Variations of the Tropical Cyclone Genesis Potential Index Based on Oceanic Parameters in the Global Ocean

PAN Lixia, WANG Xin, ZHOU Lei, and WANG Chunzai

Climatological and Seasonal Variations of the Tropical Cyclone Genesis Potential Index Based on Oceanic Parameters in the Global Ocean

PAN Lixia1), 4), WANG Xin1), 2), 3), *, ZHOU Lei5), 6), and WANG Chunzai1), 2), 3)

1) State Key Laboratory of Tropical Oceanography, South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou 510301, China 2) Innovation Academy of South China Sea Ecology and Environmental Engineering, Chinese Academy of Sciences, Guangzhou 510301, China 3) Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou), Guangzhou 511458, China 4) University of Chinese Academy of Sciences, Beijing 100049, China 5) School of Oceanography, Shanghai Jiao Tong University, Shanghai 200030, China 6) Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai 519080, China

This study investigates the global performance of the tropical cyclone (TC) genesis potential index based on oceanic parameters (GPIocean) proposed by Zhang(2016). In six major TC formation basins, GPIoceancan represent the seasonal variations of TC genesis over most basins, except for the North Indian Ocean (NIO). The monthly climatological GPIoceanshows only a single peak in the NIO, which cannot describe the bimodal pattern of the annual cycle of TC genesis. To determine the cause of the poor performance of GPIoceanin the NIO, the relative contributions of different parameters related to GPIoceanare calculated and compared with those related to the genesis potential index developed by Emanuel and Nolan (2004) (GPI04). Results show that the net longwave radiation on the sea surface is responsible for the single peak of TC genesis in the NIO in boreal summer. Compared with GPI04, vertical wind shear is not involved in GPIocean. Vertical wind shear is the dominant factor inhibiting TC genesis in the NIO in boreal summer. Therefore, the absence of vertical wind shear in GPIoceanresults in the failure of the annual cycle of TC genesis in the NIO.

North Indian Ocean; tropical cyclone; genesis potential index

1 Introduction

Tropical cyclones (TCs) are one of the most devastating natural disasters that are generated in the ocean. Because TC genesis largely depends on environmental conditions, it is of importance to understand the influence of large-scale environmental parameters on TC genesis. Gray (1967) identified six parameters that considerably influenced TC genesis, that is, i) low-level relative vorticity, ii) vertical shear of horizontal winds, iii) Coriolis parameter (at least a few degrees poleward of the equator), iv) sea surface temperature (SST) threshold (usually taken as 26℃), v) conditional instability through the air column, and vi) humidity in the lower and middle levels of the troposphere. To analyze the relationship between environmental para- meters and TC genesis quantitatively, Gray (1979) developed an empirical index for TC genesis. This index provides an empirical quantification of the relative contribu-tions of various environmental factors to TC genesis.

Emanuel and Nolan (2004) developed a genesis potential index (referred to as GPI04 hereafter) by considering four parameters related to TC genesis, that is, potential intensity (PI), relative humidity, low-troposphere wind vorticity, and vertical wind shear. GPI04 is widely applied to analyze the variations of TC activity on multiple timescales for further use in conducting reanalysis and deriving model outputs (Camargo, 2007a, 2007b; Nolan, 2007; Vecchi and Soden, 2007; Camargo, 2009; Daloz and Camargo, 2018; Zhang, 2018). For example, Camargo(2007a) used the index to diagnose the effects of the El Niño-Southern Oscillation (ENSO) on cyclone genesis and found that GPI04 could reproduce the variations of the observed frequency and location of TC genesis in the global ocean during El Niño and La Niña. Some modifications have been made to improve GPI04 (Emanuel, 2010; Murakami and Wang, 2010; McGauley and Nolan, 2011; Tippett, 2011; Bruyère, 2012; Wang and Moon, 2017). Because vertical velocity is essential for TC genesis, it was added to GPI04 by Murakami and Wang (2010). Moreover, several studies defined different indices from GPI04 (Bye and Keay, 2008; Tang and Emanuel, 2012; Waters, 2012). For example, Wa- ters(2012) considered the importance of the phase of the Madden-Julian oscillation (MJO) and equatorial wave activity to medium-frequency to high-frequency tropical cyclogenesis variability. They developed metrics for medium-frequency to high-frequency (15-day base period) variability of environmental conditions and assessed its utility as a diagnostic index for TC genesis in the North Atlantic (NATL) main development region. The details of these indices are shown in Table 1. All of these modified indices facilitate studies of the environmental effects of TC genesis and help predict TC activity.

The previously mentioned GPIs mainly considered the atmospheric parameters. The SST is usually the only parameter used to characterize ocean contribution as a heat source of TC genesis. Recently, several studies found that oceanic parameters, such as surface heat flux, mixed layer depth, upper ocean heat content, and ocean wave, rather than SST could significantly influence TC activities (Shay, 2000; Black, 2007; Wu, 2007; Price, 2009; Scoccimarro, 2011; Lee and Chen, 2012; Aijaz, 2017). Several general circulation models that combine the TC forecast system with three-dimensional ocean circulation models can make detailed forecasts of specific storms (Ginis, 2002; Bender, 2007; Chen, 2007; Dong, 2017; Balaguru, 2018). These studies showed the importance of oceanic parameters to TC activity (, intensity). Several researchers investigated the effect of oceanic factors on TC genesis according to the modified PI by changing the SST (Vechi and Soden, 2007; Lin, 2013). For example, Lin(2013) replaced the SST with the mean temperature in the pre-cyclone upper thermocline in the PI index, which improved its performance.

Given the importance of oceanic parameters, Zhang(2016) took oceanic factors into account and defined a new GPI. Various oceanic factors have been selected as candidate factors on the basis of the physical understanding of the effects of oceanic factors on TCs, and several necessary atmospheric factors are also included. To better address the roles of the ocean, they considered the effects of the atmospheric parameters on the sea surface. Four parameters were used to define the new GPI (referred to as GPIoceanhereafter), including i) absolute vorticity at 1000 hPa, ii) net longwave radiation on the sea surface, iii) mean ocean temperature in the upper mixed layer, and iv) depth of the 26℃ isotherm. Compared with previous GPIs, GPIoceanconsiders the contributions of oceanic processes that reflect the roles of subsurface factors in TC genesis. GPIoceancan reproduce TC genesis in the Western North Pacific (WNP), which shows that oceanic factors have a statistically significant relationship with TC genesis (Zhang, 2016). GPIoceanprovides a quantitative tool that connects the subsurface oceanic environment and long-term variability of TC genesis.

Table 1 Details of various indices

Although GPIoceancan represent the seasonal and interannual variations and long-term variability of TC activities in the WNP (Zhang, 2016), its performance in the global ocean is still unknown. Therefore, the present study aims to evaluate GPIoceanin the global ocean with respect to the seasonal variations of TC genesis and investigate the possible causes of the poor performance of GPIocean.

The remainder of this paper is organized as follows: Section 2 introduces the datasets and methods used in the study. Section 3 evaluates and compares the performance of GPIoceanin the global ocean with that of GPI04. Section 4 presents the conclusions and discussion.

2 Datasets and Methods

2.1 Datasets

The monthly atmospheric variables (, wind vorticity, relative humidity, and vertical pressure velocity) and surface heat flux (, shortwave radiation, longwave radiation, and latent heat flux) are obtained from the Medium Range Weather Forecasts interim reanalysis (ERA-Interim reanalysis) with a resolution of 1˚ longitude×1˚ latitude (Dee, 2011). The monthly National Center for Environmental Prediction/National Center for Atmospheric Research reanalysis (Kalnay, 1996) is used to calculate GPI04 and GPIocean, which show similar results to those from the ERA-Interim reanalysis. The upward heat flux from the ocean to the atmosphere is determined to have a positive value. The monthly SST is obtained from the Met Office Hadley Center with a resolution of 1˚ longitude×1˚ latitude (Rayner, 2003). The mean ocean temperature is the EN4 quality-controlled ocean data (EN4.0.2) derived from the Hadley Center subsurface temperature and salinity objective analyses with a horizontal resolution of 1˚ longitude×1˚ latitude and a vertical resolution of approximately 1m apart at the top 100m of the ocean, 10m apart above 1500m depth, and 50m apart below that depth (Good, 2013). Global TC genesis data are obtained from the Joint Typhoon Warning Center. The analysis period is from 1979 to 2016 in the present study.

2.2 Methods

According to Zhang(2016), GPIoceancan be calculated as follows:

,(1)

According to Emanuel and Nolan (2004), GPI04 can be calculated as follows:

where8500is the absolute vorticity at 850hPa,shearis the magnitude of vertical wind shear between 850hPa and 200hPa and is calculated using the formula

is the relative humidity at 600hPa,potis the maximum TC PI, andis set as 7.4×10−3.

To investigate the contributions of various large-scale environmental factors to GPI04 and GPIocean, the method proposed by Li(2013) can be expressed as follows:

where GPI denotes GPIoceanor GPI04,is the constant co- efficient in Eqs. (1) and (2) for GPIoceanor GPI04,

whereas

Applying the total differential to both sides of Eq. (3) yields the following expression:

Integrating Eq. (4) from the annual mean to a particular month yields the following expression:

where1,2,3, and4are assumed to be constant coefficients and expressed as follows:

where the bar denotes the annual mean value and δdenotes the difference between an individual month and the annual mean in Eq. (5). The four terms on the right side of Eq. (5) denote the contributions of each environmental factor to GPI04 or GPIocean.

3 Results

3.1 Global Distribution of GPIocean

The climatological GPIoceanand GPI04 during the period 1979–2016 are shown in Figs.1a and 1b, respectively

TC mainly occurs in six basins, namely, North Indian Ocean (NIO; 5˚ to 15˚N, 67˚ to 95˚E), WNP (5˚ to 20˚N, 130˚ to 150˚E), eastern North Pacific (ENP; 10˚ to 20˚N, 240˚ to 260˚E), NATL (10˚ to 20˚N, 310˚ to 340˚E), south- ern Indian Ocean (SIO; 5˚ to 15˚S, 55˚ to 100˚E), and southwestern Pacific (SWP; 5˚ to 15˚S, 150˚ to 180˚E) (Fig.1c). In general, GPIoceanand GPI04 have high values inall of these well-known TC-prone regions (Figs.1a and 1b). Differences between the spatial distributions of GPIoceanand GPI04 are observed. GPIoceanshows maximum values along the Kuroshio and the Gulf Stream. The amplitude of GPIoceanin the ENP is less than that in the WNP, whereas the amplitude of GPI04 in these two basins is comparable. The center of GPIoceanin the SWP shifts eastward compared with that of GPI04. These two indices also perform differently along the coast of the South American continent. GPIoceanshows a positive value along the coast of the Pacific side because no TC genesis occurs in the region, whereas GPI04 does not show a significant positive value along the coast of the Pacific side. Along the coastal regions of the Atlantic side, the significant positive value of GPIoceancan represent the generation of three TCs, which cannot be represented by GPI04.

To investigate the details of GPIocean, the annual cycles in six basins are individually examined. Consistent with the results of a previous study, the frequency of TC genesis in all basins shows a single peak, except for the NIO (Fig.2). The peaks of TC genesis in the WNP, ENP, and NATL occur in boreal summer and autumn (Figs.2b, 2c, and 2d), whereas the peaks of TC genesis in the SIO and WSP occur in boreal winter (Figs.2e and 2f). The frequency of TC genesis in the NIO has two peaks, that is, during the pre-monsoon (April–May) and post-monsoon (October–November) periods, and only a few TCs occur during the southwest monsoon period (June–September; Fig.2a). The annual cycles of GPI04 in six basins are consistent with those of TC genesis. GPIoceanshows a single peak in all basins. In the NIO, GPIoceanshows a peak in boreal summer and fails to represent the observed bimodal pattern of the annual cycle of TC genesis (Fig.2a). Therefore, we focus on the NIO.

Fig.1 Global distribution of climatological (a) GPIocean, (b) GPI04, and (c) tropical cyclone (TC) genesis during the period 1979–2016. The black dots in (c) denote the locations of TC genesis, and the blue boxes represent the six major regions of TC genesis, namely, North Indian Ocean (NIO; 5˚ to 15˚N, 67˚ to 95˚E), western North Pacific (WNP; 5˚ to 20˚N, 130˚ to 150˚E), eastern North Pacific (ENP; 10˚ to 20˚N, 240˚ to 260˚E), North Atlantic (NATL; 10˚ to 20˚N, 310˚ to 340˚E), southern Indian Ocean (SIO; 5˚ to 15˚S, 55˚ to 100˚E), and southwestern Pacific (SWP; 5˚ to 15˚S, 150˚ to 180˚E). The map shown in this figure was generated bythe National Center for Atmospheric Research (NCAR) Command Language (Version 6.6.2) [Software] (2019). Boulder, Colorado: UCAR/NCAR/CISL/TDD. http://dx.doi.org/10.5065/D6WD3XH5.

Fig.2 Annual cycle of GPIocean (blue curve), GPI04 (red curve), and TC genesis (gray bar) averaged in the (a) NIO, (b) WNP, (c) ENP, (d) NATL, (e) SIO, and (f) WSP regions during the period 1979–2016. The left y-axis denotes TC genesis, and the right y-axis denotes the GPIocean or GPI04 values.

3.2 Cause of the Poor Performance of GPIocean in the NIO During the Summer Monsoon

Fig.2a shows the failure of GPIoceamto depict the annual cycle of TC in the NIO. The spatial distributions of the climatological monthly GPIoceamin the NIO are investigated. Fig.3 shows that GPIoceanhas a high value in the northern Bay of Bengal and the Arabian Sea in boreal summer (June–September; Figs.3f to 3i). However, only a few TCs occur, particularly in July and August. TCs are usually generated in the region with high values of GPIoceamin the other months, indicating that GPIoceamcould represent the features of TC genesis in boreal autumn, winter, and spring.

To determine why GPIoceancannot represent the bimodal pattern of the annual cycle of TC genesis in the NIO, the method proposed by Li(2013) is used to identify the relative contribution of each environmental factor to GPIocean. According to Eq. (7), Fig.4a shows the climatological monthly contributions of four environmental factors of GPIoceanand the sum of these four terms. The figure illustrates that the combined contributions of the four parameters are favorable for TC genesis in May–November, which is consistent with the results shown in Fig.2a. GPIoceanshows a peak during the summer monsoon (June–August), which is mostly controlled by the positive contribution of net longwave radiation on the sea surface. How- ever, only a few TCs occur in the NIO in boreal summer. This finding indicates that several important environmental factors influencing TC genesis in the NIO are missing in GPIocean.

In contrast to GPIocean, GPI04 can represent the occurrence of only a few TCs in the NIO in boreal summer (Fig.2a), which helps in finding the missing environmental factor in GPIocean. The analysis of the individual contributions of the four parameters of GPI04 showed that a strong environmental vertical wind shear is unfavorable for TC genesis and can offset the positive contributions of relative humidity together with other environmental factors (Fig.4b). Vertical wind shear is not involved in GPIocean, thus resulting in the failure to reproduce the two peaks of TC genesis in the NIO in the seasonal cycle.

Fig.3 Climatological monthly GPIocean in the NIO. The black dots denote individual genesis events during the period 1979–2016. The map shown in this figure was generated by the NCAR Command Language (Version 6.6.2) [Software] (2019). Boulder, Colorado: UCAR/NCAR/CISL/TDD. http://dx.doi.org/10.5065/D6WD3XH5.

Fig.4 Climatological monthly contributions of each term of (a) GPIocean and (b) GPI04 in the NIO (denoted by a specified bar). The coefficients are,,, and .,,, and in (a), and ,, , and in (b). The red solid line denotes the value of δGPIocean and δGPI04 in (a) and (b), respectively.

Previous studies showed that a strong environmental vertical wind shear is unfavorable for TC genesis and is the main reason for the suppression of TC genesis in the NIO during the summer monsoon (Gray, 1967; Camargo, 2007a; Evan and Camargo, 2011). Strong vertical wind shear suppresses TC genesis by ventilating the incipient disturbance with low-entropy (low-equivalent potential temperature) air (Tang and Emanuel, 2012). The environmental flow also advects dry air into the disturbance, disrupting the formation of a deep, moist column that is postulated to be imperative for genesis (Bister and Emanuel, 1997; Nolan, 2007). Li(2013) found that the vertical wind shear cap for TC genesis in the Bay of Bengal is approximately 24ms−1. Values larger than the cap inhibit TC formation. Although the maximum value of vertical wind shear is located at approximately 50˚ to 60˚E, 10˚ to 20˚N, the value over most parts of the NIO is larger than the cap, thus inhibiting TC genesis during the summer monsoon (Fig.5). This finding indicates that vertical wind shear plays a dominant role in TC genesis in the NIO. Because of the missing vertical wind shear, GPIoceanoverestimates the TC genesis in the NIO in boreal summer; thus, it cannot reproduce the two peaks in April–May and October–December.

Fig.5 Climatological monthly vertical wind shear in the NIO from June to August during the period 1979–2016 (unit: ms−1). The solid black line denotes the isoline of 24ms−1.

4 Discussion and Conclusions

In previous studies, the atmospheric parameters are used to define GPIs to describe the spatial and temporal distri- butions of TCs. The roles of oceanic parameters in TC gen- esis are not considered, except for SST. However, more re- search showed that oceanic parameters in addition to SST play important roles in modulating TC activity (Shay, 2000; Ginis, 2002; Bender, 2007; Black, 2007; Chen, 2007; Wu, 2007; Halliwell Jr., 2008; Price, 2009; Scoccimarro, 2011; Lin, 2013). Thus, Zhang(2016) defined a new index (, GPIocean) on the basis of several oceanic parameters that significantly affect TC genesis.

The present study investigated the global distribution of GPIocean. The results show that GPIoceancould represent the annual cycle of TC genesis in the global ocean, except for the NIO. GPIoceanshows a peak during the summer monsoon in the NIO, whereas only a few TCs occur. To determine why GPIoceanfails to represent the annual cycle of TC genesis in the NIO, the relative contribution of each factor to GPIoceanis calculated and addressed on the basis of the method proposed by Li(2013). The results show that the net longwave radiation on the sea surface is responsible for the false peak of TC genesis in the NIO in boreal summer, and the three other oceanic factors do not contribute to the false peak of TC genesis. Compared with GPI04, vertical wind shear is not involved in GPIocean. Vertical wind shear in the NIO in summer is strong because of the summer monsoon, which considerably inhibits TC genesis. Therefore, the absence of vertical wind shear in GPIoceanresults in the failure of the annual cycle of TC genesis in the NIO.

Acknowledgements

This research is supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (No. XDA20060502), the National Key Research and Development Program of China (No. 2019YFA0606701), the National Natural Science Foundation of China (Nos. 41925024 and 41731173), the Pioneer Hundred Talents Program of the Chinese Academy of Sciences, the Leading Talents of Guangdong Province Program, Innovation Academy of South China Sea Ecology and Environmental Engineering, Chinese Academy of Sciences (No. ISEE2018PY06), and the Key Special Project for Introduced Talents Team of Southern Marine Science and Engineering Guangdong Lab- oratory (Guangzhou) (No. GML2019ZD0306).

Aijaz, S., Ghantous, M., Babanin, A. V., Ginis, I., Thomas, B., and Wake, G., 2017. Nonbreaking wave-induced mixing in upper ocean during tropical cyclones using coupled hurricane- ocean-wave modeling., 122: 3939-3963, https://doi.org/10.1002/2016JC012219.

Balaguru, K., Foltz, G. R., Leung, L. R., Hagos, S., and Judi, D. R., 2018. On the use of ocean dynamic temperature for hurricane intensity forecasting., 33 (2): 411-418, https://doi.org/10.1175/WAF-D-17-0143.1.

Bender, M. A., Ginis, I., Tuleya, R., Thomas, B., and Marchok, T., 2007. The operational GFDL coupled hurricane-ocean prediction system and a summary of its performance.,135: 3965-3989,https://doi.org/10.1175/2007MWR2032.1.

Bister, M., and Emanuel, K. A., 1997. The genesis of Hurricane Guillermo: TEXMEX analyses and a modeling study.,125: 2662-2682,https://doi.org/10.1175/1520-0493(1997)125<2662:TGOHGT>2.0.CO;2.

Black, P. G., Dasaro, E. A., Drennan, W. M., French, J. R., Niiler, P. P., Sanford, T. B.,, 2007. Air-sea exchange in hurri- canes: Synthesis of observations from the coupled boundary layer air-sea transfer experiment., 88 (3): 357-374, https://doi.org/10.1175/BAMS-88-3-357.

Bruyère, C. L., Holland, G. J., and Towler, E., 2012. Investigating the use of a genesis potential index for tropical cyclones in the North Atlantic Basin., 25: 8611-8626,https://doi.org/10.1175/JCLI-D-11-00619.1.

Bye, J., and Keay, K., 2008. A new hurricane index for the Car- ibbean., 33: 556-560.

Camargo, S. J., Emanuel, K. A., and Sobel, A. H., 2007a. Use of a genesis potential index to diagnose ENSO effects on tropicalcyclone genesis., 20: 4819-4834,https://doi.org/10.1175/JCLI4282.1.

Camargo, S. J., Sobel, A. H., Barnston, A. G., and Emanuel, K. A., 2007b. Tropical cyclone genesis potential index in climate models., 59A: 428-443, https://doi.org/10.1111/j.1600-0870.2007.00238.x.

Camargo, S. J., Wheeler, M. C., and Sobel, A. H., 2009. Diagno- sis of the MJO modulation of tropical cyclogenesis using an empirical index., 66: 3061-3074,https://doi.org/10.1175/2009JAS3101.1.

Chen, S. S., Price, J. F., Zhao, W., Donelan, M. A., and Walsh, E. J., 2007. The CBLAST-Hurricane program and the next-gen- eration fully coupled atmosphere-wave-ocean models for hur- ricane research and prediction., 88: 311-318.

Daloz, A. S., and Camargo, S. J., 2018. Is the poleward migration of tropical cyclone maximum intensity associated with a pole- ward migration of tropical cyclone genesis.,50 (1-2): 705-715, https://doi.org/10.1007/s00382-017-3636-7.

Dee, D., Uppala, S., Simmons, A. J., Berrisford, P., Poli, P., Kobayashi, S.,, 2011. The ERA-interim reanalysis: Con- figuration and performance of the data assimilation system., 137 (656): 553-597, https://doi.org/10.1002/qj.828.

Dong, J., Domingues, R., Goni, G., Halliwell, G. R., Kim, H., Lee, S.,, 2017. Impact of assimilating underwater glider data on Hurricane Gonzalo (2014) forecasts., 32 (3): 1143-1159, https://doi.org/10.1175/WAF- D-16-0182.1.

Emanuel, K., 2010. Tropical cyclone activity downscaled from NOAA-CIRES reanalysis, 1908–1958.,2: 1-12, https://doi.org/10.3894/JAMES.2010.2.1.

Emanuel, K., and Nolan, D. S., 2004. Tropical cyclone activity and the global climate system.. Miami, 240-241.

Evan, A. T., and Camargo, S. J., 2011. A climatology of Arabian Sea cyclonic storms., 24: 140-158,https://doi.org/10.1175/2010JCLI3611.1.

Ginis, I., 2002. Tropical cyclone-ocean interactions., 33: 83-114,https://doi.org/10.1145/3093338.3104149.

Good, S. A., Martin, M. J., and Rayner, N. A., 2013. EN4: Quali- ty controlled ocean temperature and salinity profiles and month- ly objective analyses with uncertainty estimates.,118: 6704-6716,https://doi.org/10.1002/2013JC009067.

Gray, W. M., 1967. Global view of the origin of tropical distur- bances and storms.,96 (10): 669-700,https://doi.org/10.1175/1520-0493(1968)096<0669:GVOTOO >2.0.CO;2.

Gray, W. M., 1979. Hurricanes: Their formation, structure and likely role in the tropical circulation., 77: 155-218.

Halliwell Jr., G. R., Shay, L. K., Jacob, S. D., Smedstad, O. M., and Uhlhorn, E. W., 2008. Improving ocean model initialization for coupled tropical cyclone forecast models using GOD- AE nowcasts., 136 (7): 2576-2591, https://doi.org/10.1175/2007MWR2154.1.

Kalnay, E., Kanamitsu, M., Kistler, R., Collins, W. D., Deaven, D. G., Gandin, L. S.,, 1996. The NCEP/NCAR 40-year reanalysis project., 77: 437-472,https://doi.org/10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

Lee, C.-Y., and Chen, S. S., 2012. Symmetric and asymmetric structures of hurricane boundary layer in coupled atmosphere-wave-ocean models and observations., 69 (12): 3576-3594, https://doi.org/10.1175/JAS-D-12-046.1.

Li, Z., Yu, W., Li, T., Murty, V. S. N., and Tangang, F., 2013. Bimodal character of cyclone climatology in the Bay of Bengal modulated by monsoon seasonal cycle., 26: 1033-1046, https://doi.org/10.1175/JCLI-D-11-00627.1.

Lin, I. I., Black, P. G., Price, J. F., Yang, C., Chen, S. S., Lien, C.,, 2013. An ocean coupling potential intensity index for tropical cyclones., 40: 1878-1882,https://doi.org/10.1002/grl.50091.

McGauley, M. G., and Nolan, D. S., 2011. Measuring environ- mental favorability for tropical cyclogenesis by statistical anal- ysis of threshold parameters., 24: 5968-5997,https://doi.org/10.1175/2011JCLI4176.1.

Murakami, H., and Wang, B., 2010. Future change of North Atlantic tropical cyclone tracks: Projection by a 20-km-mesh global atmospheric model., 23: 2699-2721,https://doi.org/10.1175/2010JCLI3338.1.

Nolan, D. S., 2007. What is the trigger for tropical cyclogenesis?, 56 (4): 241-266.

Nolan, D. S., Rappin, E. D., and Emanuel, K. A., 2007. Tropical cyclogenesis sensitivity to environmental parameters in radi- ative-convective equilibrium.,133: 2085-2107,https://doi.org/10.1002/qj.170.

Price, J. F., 2009. Metrics of hurricane-ocean interaction: Ver- tically-integrated or vertically-averaged ocean temperature?, 5: 351-368, https://doi.org/10.5194/os-5-351-2009.

Rayner, N. A., Parker, D. E., Horton, E. B., Folland, C. K., Alexander, L. V., Rowell, D. P.,, 2003. Global analyses of sea surface temperature, sea ice, and night marine air tem- perature since the late nineteenth century., 108 (D14): 4407, https://doi.org/10.1029/2002JD002670.

Scoccimarro, E., Gualdi, S., Bellucci, A., Sanna, A., Fogli, P. G., Manzini, E.,, 2011. Effects of tropical cyclones on ocean heat transport in a high-resolution coupled general circulation model., 24 (16): 4368-4384, https://doi.org/10.1175/2011JCLI4104.1.

Shay, L. K., Goni, G. J., and Black, P. G., 2000. Effects of a warm oceanic feature on Hurricane Opal.,128: 1366-1383, https://doi.org/10.1175/1520-0493(2000)128<1366:EOAWOF>2.0.CO;2.

Tang, B., and Emanuel, K., 2012. A ventilation index for tropical cyclones., 93: 1901-1912,https://doi.org/10.1175/BAMS-D-11-00165.1.

Tippett, M. K., Camargo, S. J., and Sobel, A. H., 2011. A Poisson regression index for tropical cyclone genesis and the role of large-scale vorticity in genesis., 24: 2335-2357,https://doi.org/10.1175/2010JCLI3811.1.

Vecchi, G. A., and Soden, B. J., 2007. Increased tropical Atlantic wind shear in model projections of global warming., 34: L08702, https://doi.org/10.1029/2006GL028905.

Wang, B., and Moon, J., 2017. An anomalous genesis potential index for MJO modulation of tropical cyclones., 30 (11): 4021-4035,https://doi.org/10.1175/jcli-d-16-0749.1.

Waters, J. J., Evans, J. L., and Forest, C. E., 2012. Large-scale diagnostics of tropical cyclogenesis potential using environ- ment variability metrics and logistic regression models.,25: 6092-6107,https://doi.org/10.1175/JCLI-D-11-00359.1.

Wu, C. C., Lee, C. Y., and Lin, I. I., 2007. The effect of the ocean eddy on tropical cyclone intensity.,64: 3562-3578,https://doi.org/10.1175/JAS4051.1.

Zhang, M., Zhou, L., Chen, D., and Wang, C., 2016. A genesis potential index for western North Pacific tropical cyclones by using oceanic parameters.,121: 6762-6778,https://doi.org/10.1002/2016JC011851.

Zhang, W., Villarini, G., Vecchi, G. A., and Murakami, H., 2018. Impacts of the Pacific Meridional Mode on landfalling North Atlantic tropical cyclones., 50 (3): 991-1006,https://doi.org/10.1007/s00382-017-3656-3.

June 17, 2020;

July 7, 2020;

July 19, 2020

© Ocean University of China, Science Press and Springer-Verlag GmbH Germany 2021

. Tel: 0086-20-34134191 E-mail: wangxin@scsio.ac.cn

(Edited by Xie Jun)