The tropical cyclone records were extracted from the International Best Track Archive for Climate Stewardship (IBTrACS; Knapp et al., 2010Knapp, K. R.; Kruk, M. C.; Levinson, D. H.; Diamond, H. J. & Neumann, C. J. 2010. “The International Best Track Archive for Climate Stewardship (IBTrACS): Unifying tropical cyclone best track data”. Bulletin of the American Meteorological Society, 91: 363-376, ISSN: 1520-0477, DOI:10.1175/2009BAMS2755.1. ), which is freely available at https://www.ncdc.noaa.gov/ibtracs/index.php?name=climatology. The IBTrACS is officially recognized by the WMO Tropical Cyclone Programme as an official TC data resource and is under the auspices of the World Data Center for Meteorology located at NOAA’s National Centers for Environmental Information. It widely discuss by several authors (e.g. Bathia et al., 2019Bhatia, K. T.; Vecchi, G. A.; Knutson, T. R.; Murakami, H.; Kossin, J.; Dixon, K. W. & Whitlock, C. E. 2019. “Recent increases in tropical cyclone intensification rates”. Nature Communication, 10: 635, ISSN 2041-1723, DOI: 10.1038/s41467-019-08471-z.; Kossin et al., 2013Kossin, J. P.; Olander, T. L. & Knapp, K. R. 2013. “Trend Analysis with a New Global Record of Tropical Cyclone Intensity”. Journal of Climate, 26; 9960-9976, ISSN: 0894-8755, DOI: 10.1175/JCLI-D-13-00262.1.; Vecchi and Knutson, 2008Vecchi, G. A. & Knutson, T. R. 2008. “On Estimates of Historical North Atlantic Tropical Cyclone Activity”. Journal of Climate, 21(14): 3580-3600, ISSN: 0894-8755, DOI: 10.1175/2008JCLI2178.1.; Chang and Guo, 2007Chang, E. K. M. & Guo, Y. 2007. “Is the number of North Atlantic tropical cyclones significantly underestimated prior to the availability of satellite observations?”. Geophysical Research Letter, 34: L14801, ISSN: 1944-8007, DOI: 10.1029/2007GL030169.) that before the era of air reconnaissance flights and meteorological satellites, the detection of TCs depended on fortuitous encounters with ships or their impact on populated areas. This means that the reliability of the long-term hurricane record is dependent on who was measuring them, and how, at any given time. Therefore, the quality of the satellite images made possible to determine the location and classification of the TCs more reliable after the 1970s. In agreement with Bathia et al., 2019Bhatia, K. T.; Vecchi, G. A.; Knutson, T. R.; Murakami, H.; Kossin, J.; Dixon, K. W. & Whitlock, C. E. 2019. “Recent increases in tropical cyclone intensification rates”. Nature Communication, 10: 635, ISSN 2041-1723, DOI: 10.1038/s41467-019-08471-z., in this study were used the tropical cyclone records from 1980 to 2019. Also, two different statistical tests - The Power of Pruned Exact Linear Time (PELT) (Killick et al., 2012Killick, R.; Fearnhead, P. & Eckley, I. A. 2012. “Optimal detection of change points with a linear computational cost”. Journal of the American Statistical Association, 107(500): 1590-1598, ISSN: 1537-274X, DOI: 10.1080/01621459.2012.737745.) and Binary Segmentation (BINSEG) (Scott and Knott, 1974Scott, A. J. & Knott, M. 1974. “A Cluster Analysis Method for Grouping Means in the Analysis of Variance”. Biometrics, 30(3): 507-512, ISSN: 0006341X.)- were applied to determine the change points in this time series and corroborate our selection.

From the Centennial Time Scale (COBE SST2) dataset (Hirahara et al., 2014Hirahara, S.; Ishii, M. & Fukuda, Y. 2014 “Centennial-scale sea surface temperature analysis and its uncertainty”. Journal of Climate, 27: 57-75, ISSN: 0894-8755, DOI: 10.1175/JCLI-D-12-00837.1.) of the National Oceanic and Atmospheric Administration (NOAA), freely available at https://psl.noaa.gov/data/gridded/data.cobe2.html, was obtained the monthly mean SST data. The COBE SST is a monthly mean of global fields created during June, 2011 at NOAA’s Physical Sciences Division (PSD) of the Earth System Research Laboratory (ESRL) using the gridded data from the Japanese Reanalysis Project (JRA). Moreover, all the climate variability modes used here were extracted from the NOAA Physical Sciences Laboratory at https://psl.noaa.gov/data/climateindices/list/. The NAO consists of a north-south dipole of anomalies, with a centre located over Greenland and the other centre of opposite sign that covers the central latitudes of the NATL between 35°N and 40°N. The NAO index is defined as the normalized pressure difference between the Azores and Iceland (Hurrell, 1995Hurrell, J. W. 1995. “Decadal trends in the North Atlantic Oscillation and relationships to regional temperature and precipitation”. Science, 269: 676-679, ISSN: 1095-9203.; Jones et al., 1997Jones, P. D.; Jónsson, T. & Wheeler, D. 1997. “Extension to the North Atlantic Oscillation using early instrumental pressure observations from Gibraltar and South-West Iceland”. International Journal of Climatology, 17: 1433-1450, ISSN: 1097-0088, DOI: 10.1002/(SICI)1097-0088(19971115)17:13<1433::AID-JOC203>3.0.CO;2-P.). The AMO is a climate variability mode that also occurs in the NATL; its index is calculated as the area weighted average SST over the North Atlantic, basically between 0° to 70°N (Enfield et al., 2001Enfield, D. B.; Mestas-Nunez, A. M. & Trimble, P. J. 2001. “The Atlantic Multidecadal Oscillation and its relationship to rainfall and river flows in the continental U.S”. Geophysical Research Letter,28: 2077-2080, ISSN: 1944-8007, DOI: 10.1029/2000GL012745 ). On the other hand, the AMM mode describes the meridional variability in the tropical Atlantic Ocean applying Maximum Covariance Analysis (MCA) to the SST and the zonal and meridional components of the 10m wind field (Chiang and Vimont, 2004Chiang, J. C. H. & Vimont, D. J. 2004. “Analagous meridional modes of atmosphere-ocean variability in the tropical Pacific and tropical Atlantic”. Journal of Climate, 17(21): 4143-4158, ISSN: 0894-8755, DOI: 10.1175/JCLI4953.1.). The Caribbean index (CAR) is the SST anomalies averaged over the Caribbean Sea, and the NTA index is the SST anomalies averaged over 60°-20°W and 6°-18°N, and 20°-10°W, 6°-10°N. Both indices were calculated relative to the 1981-2010 climatology and smoothed by three months running mean procedure (Penland and Matrosova, 1998Penland, C. & Matrosova, L. 1998. “Prediction of tropical Atlantic sea surface temperatures using Linear Inverse Modeling”. Journal of Climate, 11(3): 483-496, ISSN: 0894-8755, DOI: 10.1175/1520-0442(1998)011<0483:POTASS>2.0.CO;2 ). The quasi-biennial oscillation (QBO) is a quasi-periodic equatorial zonal wind oscillation between the east and west winds in the tropical stratosphere with a mean period of 28 to 29 months. It is calculated from the zonal average of the 30mb zonal wind at the equator (Andrews et al., 1987Andrews, D. G.; Holton, J. R. & Leovy, C. B. 1987. Middle Atmosphere Dynamics. 1st ed., vol. 40, United Kingdom: Academic Press, 489p., ISBN: 9780080511672, Available: <https://www.sciencedirect.com/bookseries/international-geophysics/vol/40/suppl/C>, [Consulted: Febraury 10, 2021].; Naujokat, 1986Naujokat, B. 1986. “An update of the observed quasi-biennial oscillation of the stratospheric winds over the tropics”. Journal of Atmospheric Science, 43: 1873-1877, ISSN: 1520-0469, DOI: 10.1175/1520-0469(1986)043<1873:AUOTOQ>2.0.CO;2 ). The Tropical Northern Atlantic (TNA) and Tropical Southern Atlantic (TSA) indices are the anomalies of the average of the monthly SST respect to climatology 1971-2000 from 5.5°-23.5°N and 15°-57.5°W, and from 0°-20°S and 10°E-30°W, receptively (Enfield, 1999Enfield, D. B.; Mestas, A.M.; Mayer, D. A. & Cid-Serrano, L. 1999. “How ubiquitous is the dipole relationship in tropical Atlantic sea surface temperatures?”. Journal of Geophysical Research Ocean, 104: 7841-7848, ISSN: 2169-9291, DOI: 10.1029/1998JC900109.). The Bivariate ENSO index (Smith et al., 2000Smith, C. A. & Sardeshmukh, P. .2000. “The Effect of ENSO on the Intraseasonal Variance of Surface Temperature in Winter”. International Journal of Climatology, 20: 1543-1557, ISSN: 1097-0088, DOI: 10.1002/1097-0088(20001115)20:13<1543::AID-JOC579>3.0.CO;2-A.) is used to represent the ENSO conditions combining the traditional Niño 3.4 SST and the Southern Oscillation Index (SOI).

First, from the historical record was separated the frequency of TC activity by each TC intensity (from tropical depression to category 5 hurricane). To estimate where TCs were frequently generated, where intensity changes occur including hurricane intensity changes according to the Saffir-Simpson scale (Saffir,1973Saffir, H. S. 1973. “Hurricane wind and storm surge”. Military Engineering, 65(423): 4-5, ISSN: 00263982.; Simpson, 1974Simpson, R. H. 1974. “The hurricane disaster-potential scale”. Weatherwise, 27: 169-186, ISSN: 1940-1310, DOI: 10.1080/00431672.1974.9931702 ), where maximum intensity was reached by each TC, and their track paths, the nonparametric kernel density estimation (KDE) was applied (Loader, 1999Loader, C. R. 1999. “Bandwidth Selection: Classical or Plug-In?” The Annals of Statistics, 27(2): 415-438, ISSN: 00905364.). An important limitation of the KDE application is that it allows the genesis or TCs intensification over land. In this article, for the statistical analyses, the points over land were eliminated according to the grid point of SST. Additionally, the statistical t-test was applied to determine the significance of the Pearson correlations at the 95% significance level.

The application of the PELT (Killick et al., 2012Killick, R.; Fearnhead, P. & Eckley, I. A. 2012. “Optimal detection of change points with a linear computational cost”. Journal of the American Statistical Association, 107(500): 1590-1598, ISSN: 1537-274X, DOI: 10.1080/01621459.2012.737745.) and BINSEG (Scott and Knott, 1974Scott, A. J. & Knott, M. 1974. “A Cluster Analysis Method for Grouping Means in the Analysis of Variance”. Biometrics, 30(3): 507-512, ISSN: 0006341X.) methods identified significant change points in the frequency of TC genesis in the historical records in the NATL basin in 1930, 1965, 1980, and 2000 (Figure 1). The change points observed before 1980 are not relevant for this study. The 1980 change point is a consequence of changes in data collection techniques, specifically the improvements in observation methods based on satellite products and the change point detected in 2000 (Figure 2) divides the study period in two 20-year subperiods, 1980-1999 (TCsP1) and 2000-2019 (TCsP2). The whole period from 1980 to 2019 shows a slightly increasing trend in the frequency of TCs; however, both TCsP1 and TCsP2 show a decreasing trend. All trends are not statistically significant, however. The TCsP1 and TCsP2 periods exhibit an average of 14 and 17 TCs genesis events, respectively. The 1981 season was the most active of the TCsP1 period with 22 TCs, followed by less TC activity in 1982 (eight TCs) and 1983 (six TCs). Similarly, in the TCsP2 period, the 2005 season was the most active with 31 TCs followed by a less-active season in 2006 with 10 TCs.

The monthly variation in the number of genesis events and TCs that reached each category shows a maximum in September, for all cases, followed by a secondary maximum in August (Figure 2). Between both months, the 57.6% of TCs genesis is observed. This can be directly linked to SST, as both August and September are the warmest months of the hurricane season (Wang et al., 2017Wang, X.; Liu, H. & Foltz, G. R. 2017. “Persistent influence of tropical North Atlantic wintertime sea surface temperature on the subsequent Atlantic hurricane season”. Geophysical Research Letter, 44: 7927- 7935, ISSN: 1944-8007 , DOI: 10.1002/2017GL074801. ). Outside the NATL TC season, one tropical storm was formed in January, three in April, 12 in May, and six in December. Furthermore, the most active season was 2005 (not include here the 2020 TC season), with 31 TCs (including tropical depressions). During this season, the highest historical number of tropical storms (27), category 1 hurricanes (15), category 3 hurricanes (7), and category 5 hurricanes (4) were formed; the highest historical number of category 2 hurricanes (9) was recorded during the 2010 season. Additionally, the highest number of category 4 hurricanes with 5 storms occurred during both the 2005 and 2010 seasons. The mean lifetime of TCs in the NATL basin was 6.7 days and the mean maximum intensity was 89.3 km/h.

Figure 3 shows an increasing trend in the number of TCs in the NATL basin of approximately 0.8 TCs/decade. However, this trend was only statistically significant for those TCs that developed into tropical storms (TS) on the Saffir-Simpson wind scale (p < 0.05) with an increase of approximately 1.4 TS/decade from 1980 to 2019 (note that each TC category is referred as the maximum category reached by a TC in its lifetime). This is further evidence of the continued increase in cyclonic activity in the NATL basin during this period. A negligible trend was observed for hurricane category. These results support the findings of Murakami et al. (2014)Murakami, H.; Li, T. & Hsu, P. 2014. “Contributing Factors to the Recent High Level of Accumulated Cyclone Energy (ACE) and Power Dissipation Index (PDI) in the North Atlantic”. Journal of Climate, 27: 3023-3034, ISSN: 0894-8755, DOI: 10.1175/JCLI-D-13-00394.1., who state that TC activity in the NATL basin has shown a marked positive anomaly in the number of TC genesis events, their mean lifespan, and their mean maximum intensity.

Figure 4a shows the seasonal distribution of TC genesis using the classical kernel density estimation. The highest density is observed in the tropical NATL between the Lesser Antilles Arc and the West Africa coast, in agreement with DeMaria et al. (2001) DeMaria, M.; Knaff, J. A. & Connell, B. H. 2001. “A Tropical Cyclone Genesis Parameter for the Tropical Atlantic”. Weather and Forecasting, 16: 219-233, ISSN: 1520-0434, DOI: 10.1175/1520-0434(2001)016<0219:ATCGPF>2.0.CO;2 and Neumann (1993)Neumann, C. J. 1993. Global climatology. Global Guide to Tropical Cyclone Forecasting, (ser. WMO/TD No. 560, Rep. TCP-31), Technical Document, Ginebra: World Meteorological Organization. Available: <https://library.wmo.int/index.php?lvl=notice_display&id=305#.YR_jPVuxXeM>, [Consulted: Febraury 15, 2021].. In this region, the necessary conditions for barotropic-baroclinic instability are fulfilled, which acts as a source of energy for tropical waves (Molinari et al., 1997Molinari, J.; Knight, D.; Dickinson, M.; Vollaro, D. & Skubis, S. 1997. “Potential vorticity, easterly waves, and eastern Pacific tropical cyclogenesis”. Monthly Weather Review, 125: 2699-2708, ISSN: 1520-0493, DOI: 10.1175/1520-0493(1997)125<2699:PVEWAE>2.0.CO;2.). Another region with higher density values is the area between the Western Caribbean Sea, the Yucatan Peninsula, and the Bay of Campeche, associated with the reversal potential vorticity gradient (Molinari et al., 1997Molinari, J.; Knight, D.; Dickinson, M.; Vollaro, D. & Skubis, S. 1997. “Potential vorticity, easterly waves, and eastern Pacific tropical cyclogenesis”. Monthly Weather Review, 125: 2699-2708, ISSN: 1520-0493, DOI: 10.1175/1520-0493(1997)125<2699:PVEWAE>2.0.CO;2.).

The trajectory density (Figure 4b) shows the highest values in the Gulf of Mexico and northeast of the Florida Peninsula, which are associated with the tracks of the systems formed within the region. In comparison, systems formed near the West African coast show more dispersed trajectories. In this area, TCs that move towards the Caribbean Sea are as frequent as those that move directly towards the central NATL. Similar to the behaviour observed for the genesis density, similar areas of the tropical North Atlantic and the Gulf of Mexico show the highest density values where TCs reach tropical storm status (Figure 4c). Furthermore, the regions where TCs reach hurricane status have the highest density values in the Caribbean Sea, the Gulf of Mexico, the tropical NATL region at the east of the Lesser Antilles Arc, and the northeast of the Bahamas archipelago (Figures 4 d-f). The tropical NATL (below 20° latitude) and the Gulf of Mexico show the highest density values, where TCs achieve maximum intensity (Figure 4g).

Furthermore, Figure 5 shows a slight poleward migration of the mean latitude where TCs reached maximum intensity, with a not statistically significant trend of 0.17° latitude per decade This confirms the findings of Kossin et al. (2014)Kossin, J.; Emanuel, K. & Vecchi, G. 2014. “The poleward migration of the location of tropical cyclone maximum intensity”. Nature, 509: 349-352, ISSN: 1476-4687, DOI: 10.1038/nature13278., who found that the HURDAT2 and ADT-HURSAT (Kossin et al., 2013Kossin, J. P.; Olander, T. L. & Knapp, K. R. 2013. “Trend Analysis with a New Global Record of Tropical Cyclone Intensity”. Journal of Climate, 26; 9960-9976, ISSN: 0894-8755, DOI: 10.1175/JCLI-D-13-00262.1.) datasets both show no clear trend in the NATL basin for the period 1982-2009. In addition, an increase in storm track density is observed in the eastern regions of the NATL basin, which explains the decrease in track density near the Lesser Antilles Arc as shown in Figure 6. This behaviour can be attributed to changes in the steering flow of the tracks, associated with the variability of the NASH ridge.

The annual average SST (from June to November) in the NATL basin from 1980 to 2019 are shown in Figure 7. This confirms a significant warming trend of approximately 0.23 °C/decade (p < 0.05), which is similar to the trend reported by Taboada and Anadón [66] for the period 1982-2010 (0.25 °C/decade). The abrupt increase in ocean temperature since the 1970s has also been attributed to the effects of anthropogenic and natural forcing. It is notable that after (before) 2000 predominate positive (negative) SST anomalies during the hurricane season.

Figure 8 shows mean SST (computed between 5°-50°N y 10°-100°W) for the TC season in the NATL basin. No visual differences in the SST pattern can be seen; however, the 27 °C isotherm is shifted to the east in the last decade (not shown) relative to 1980-1999. The mean SST values from 1980 to 2019 show that the Gulf of Mexico was the warmest region with values higher than 29 °C from July to September; the warmest region in October and November was south of this, in the Caribbean Sea. In general, SST is higher than 27 °C in the Main Development Region of TCs in the NATL basin (MDR; 10°-25°N, 80°-20°W, Caron et al., 2015Caron, L.; Boudreault, M. & Bruyère, C. L. 2015. “Changes in large-scale controls of Atlantic tropical cyclone activity with the phases of the Atlantic multidecadal oscillation”. Climate Dynamics, 44: 1801-1821, ISSN: 1432-0894, DOI: 10.1007/s00382-014-2186-5.), which favours their genesis and development. Generally, high SST values favour evaporation over the sea surface, which leads to an intense upward moisture transport due to persistent convection, phase changes, and the consequent release of latent heat, all of which are necessary for the genesis and intensification of TCs. Positive and negative SST anomalies are observed after and before 2000, respectively, for all months of the TC season. This supports the findings of Dare and McBride (2011)Dare, R. A. & McBride, J. L. 2011. “The threshold sea surface temperature condition for tropical cyclogenesis”. Journal of Climate, 24: 4570-4576, ISSN: 0894-8755, DOI: 10.1175/JCLI-D-10-05006.1. who pointed out that SST during TC genesis were warmer after 1995.

The spatial analysis of monthly SST also indicates warming in the NATL over the last 40 years (Figure 9). This is reflected in alterations in oceanic circulation, as evidenced by modifications in energy and mass fluxes (Hakkinen et al., 2004Hakkinen, S. & Rhines, P. B. 2004. “Decline of subpolar North Atlantic gyre circulation during the 1990s”. Science, 304: 555-559, ISSN: 1095-9203, DOI: 10.1126/science.1094917.; Hakkinen et al., 2004Hakkinen, S. & Rhines, P. B. 2004. “Decline of subpolar North Atlantic gyre circulation during the 1990s”. Science, 304: 555-559, ISSN: 1095-9203, DOI: 10.1126/science.1094917.; Toggweiler, 2008Toggweiler, J. R. & Russell, J. 2008. “Ocean circulation in a warming climate”. Nature, 451: 286-288, ISSN: 1476-4687, DOI: 10.1038/nature06590.) that create more favourable conditions for the formation and intensification of TCs. Cione and Uhlhorn (2003)Cione, J. J. & Uhlhorn, E.W. 2003. “Sea Surface Temperature Variability in Hurricanes: Implications with Respect to Intensity Change”. Monthly Weather Review, 131(8): 1783-1796, ISSN: 1520-0493, DOI: 10.1175//2562.1. showed that changes in SST are directly related to changes in TC intensity as the ocean provides the necessary energy to establish and maintain deep convection. Therefore, more favourable conditions are expected for the intensification of TCs in areas with higher SSTs. Moreover, Xu et al. (2016)Xu, J.; Wang, Y. & Tan, Z. 2016. “The Relationship between Sea Surface Temperature and Maximum Intensification Rate of Tropical Cyclones in the North Atlantic”. Journal of Atmospheric Sciences, 73: 4979-4988, ISSN: 1520-0469, DOI: 10.1175/JAS-D-16-0164.1. observed a nonlinear increasing trend in the maximum potential intensification rate of TCs with increasing SSTs. They also found that the TC intensification occurs most frequently in regions where SSTs are higher than 27 °C.

Another important result shown in Figure 9 is the significant increasing SST trend above 0.3 °C/decade north of 50°N (not shown), which can lead to ice melt in the Arctic and the consequent changes in the oceanic thermohaline circulation. This has direct implications for SST anomalies (Mendelsohn et al., 2012Mendelsohn, R.; Emanuel, K. A.; Chonabayashi, S. & Bakkensen, L. 2012. “The impact of climate change on global tropical cyclone damage”. Nature Climate Change, 2: 205-209, ISSN: 1758-6798, DOI: 10.1038/nclimate1357.) and implies a consequent impact on TC activity over the NATL basin, although more in-depth studies are needed to explore this further.

Figure 10 shows monthly mean SST anomalies, the number of TCs, and the TC count for different hurricane categories. A clear transition from negative to positive anomalies is observed; negative values occur before 1995 and positive anomalies occur after 2000. Similar patterns can be seen in the number and category of TCs, with positive anomalies in 1996, 2005, 2010, and 2017.

The increase in TC activity in recent years may be directly linked to the predominant positive SST anomalies after 2000. These results seem to be related to the impact of a warming climate; however, simulations with high-resolution global models tend to indicate a decrease in the frequency of TCs under a warmer climate (Knutson et al., 2015Knutson, T. R.; Sirutis, J. J.; Zhao, M.; Tuleya, R. E.; Bender, M.; Vecchi, G. A.; Villarini, G. & Chavas, D. 2015. “Global projections of intense tropical cyclone activity for the late twenty-first century from dynamical downscaling of CMIP5/RCP4.5 scenarios”. Journal of Climate, 28(18): 7203-7224, ISSN: 0894-8755, DOI: 10.1175/jcli-d-15-0129.1.; Park and Latif, 2005Park, W. & Latif, M. 2005. “Ocean dynamics and the nature of air-sea interactions over the North Atlantic at decadal timescales”. Journal of Climate , 18: 982-95, ISSN: 0894-8755, DOI: 10.1175/JCLI-3307.1 ), associated with a saturation deficit due to higher SSTs and broadly constant relative humidity (Wehner et al., 2015Wehner, M.; Prabhat; Reed, K. A.; Stone, D.; Collins, W. D. & Bacmeister, J. 2015. “Resolution Dependence of Future Tropical Cyclone Projections of CAM5.1 in the U.S. CLIVAR Hurricane Working Group Idealized Configurations”. Journal of Climate, 28: 3905-3925, ISSN: 0894-8755, DOI: 10.1175/JCLI-D-14-00311.1.). However, these arguments have not been adequately developed and tested (Held and Soden, 2006Held, I. M. & Soden, B. J. 2006. “Robust responses of the hydrological cycle to global warming”. Journal of Climate, 19: 5686-5699, ISSN: 0894-8755, DOI: 10.1175/JCLI3990.1.), and projections of a decrease in the frequency of TCs with warming are considerably more uncertain than projected increases (Emanuel, 2013Emanuel, K. A. 2013. “Downscaling CMIP5 climate models shows increased tropical cyclone activity over the 21st century”. Proceedings of the National Academy of Sciences , 110: 12219-12224, ISSN: 1091-6490, DOI: 10.1073/pnas.1301293110.).

TC activity is influenced by several large-scale modes of climatic variability that strongly modulate seasonal patterns over different timescales (Neumann, 1993Neumann, C. J. 1993. Global climatology. Global Guide to Tropical Cyclone Forecasting, (ser. WMO/TD No. 560, Rep. TCP-31), Technical Document, Ginebra: World Meteorological Organization. Available: <https://library.wmo.int/index.php?lvl=notice_display&id=305#.YR_jPVuxXeM>, [Consulted: Febraury 15, 2021].). During the period with strong positive TC genesis anomalies (i.e., the 2005 season), many of the climate modes were in a positive phase (i.e., AMO, TNA, AMM, TSA, CAR, and NTA) combined with the negative phase of NAO and a neutral ENSO phase (Figure 11). The possible impact of mode variability on TC activity in the NATL occurs during the 2010 season, when the NAO and ENSO were in their negative phase and all other indices were in their positive phase; this was associated with the formation of 21 TCs. Notably, 2005 and 2010 were also dominated by positive SST anomalies (Figure 11).

In comparison, in 1983, when only six TCs occurred, most climate modes were in their negative phase, and the ENSO, NAO, and AMM were in the following modes: the ENSO was in a weak positive in July-August and a negative phase in September-November; the AMM was in a positive phase in June-July and November, and a negative phase in August-October; and the NAO was in a positive phase in June-August and October, and a negative phase in September and November. In 2013, all indices were positive with the exception of a weakly negative ENSO. A positive NAO induces positive anomalies in the MSLP leading to moderate vertical wind shear over the tropical NATL. These results suggest that the NAO and the AMM might modify and even oppose known ENSO impacts, which is in agreement with Lim et al. (2016)Lim, Y.; Schubert, S. D.; Reale, O.; Molod, A. M.; Suarez, M. J. & Auer, B. M. 2016. “Large-Scale Controls on Atlantic Tropical Cyclone Activity on Seasonal Time Scales”. Journal of Climate, 29: 6727-6749, ISSN: 0894-8755, DOI: 10.1175/JCLI-D-16-0098.1..

Pearson’s correlations between the different TC activity measures and the NAO were not significant (Table I). The strongest positive correlation coefficients were obtained for TC genesis and the AMO, AMM, TNA, and a negative correlation was obtained for the ENSO (r = 0.19); the negative phases of the QBO, NAO, and TSA are weakly correlated with TC genesis (r ranging from 0.13 to 0.3; Table I); and the ENSO is also inversely correlated with category 5 (r = -0.11) and category 1 (r = -0.23) hurricanes.

It is known that TCs exhibit a quasi-linear response to the ENSO in the NATL basin (Krishnamurthy et al, 2016Krishnamurthy, L.; Vecchi, G. A.; Msadek, R.; Murakami, H.; Wittenberg, A. & Zeng, F. 2016. “Impact of Strong ENSO on Regional Tropical Cyclone Activity in a High-Resolution Climate Model in the North Pacific and North Atlantic Oceans”. Journal of Climate, 29: 2375-2394, ISSN: 0894-8755, DOI: 10.1175/JCLI-D-15-0468.1.). The inhibition of TC activity during positive ENSO phases is attributed to increases in vertical wind shear and reductions in atmospheric humidity (Gray, 1984Gray, W. M. 1984. “Atlantic seasonal hurricane frequency. Part I: El Niño and 30 mb quasi-biennial oscillation influences”. Monthly Weather Review, 112(9): 1649-1668, ISSN: 1520-0493, DOI: 10.1175/1520-0493(1984)112<1649:ASHFPI>2.0.CO;2.; Lin et al., 2020Lin, I. ‐I.; Camargo, S. J.; Patricola, C. M.; Boucharel, J.; Chand, S.; Klotzbach, P.; Chan, J. C. L.; Wang, B.; Chang, P.; Li, T. & Jin, F. F. 2020. ENSO and Tropical Cyclones. In McPhaden, M. J.; Santoso, A. & Cai, W. (eds). El Niño Southern Oscillation in a Changing Climate. United States of America: American Geophysical Union (AGU), ISBN: 9781119548164, DOI: 10.1002/9781119548164.ch17.; Camargo et al., 2007Camargo, S. J.; Emanuel, K. A. & Sobel, A. H. 2007. “Use of a Genesis Potential Index to Diagnose ENSO Effects on Tropical Cyclone Genesis”. Journal of Climate, 20: 4819-4834, ISSN: 0894-8755, DOI: 10.1175/JCLI4282.1. ) Furthermore, a positive ENSO induces tropospheric warming and disequilibrium with SST, creating unfavourable conditions for TC genesis (Tang and Neelin, 2004Tang, B. H. & Neelin, J. D. 2004. “ENSO influence on Atlantic hurricanes via tropospheric warming”. Geophysical Research Letter, 31: L24204, ISSN: 1944-8007, DOI: 10.1029/2004GL021072.). An opposite pattern is observed during negative ENSO phases (Deser et al., 2010Deser, C.; Alexander, M. A.; Xie, S.-P. & Phillips, A. S. 2010. “Sea surface temperature variability: Patterns and mechanisms”. Annual Review of Marine Science, 2: 115-143, ISSN: 1941-0611, DOI:10.1146/annurev-marine-120408-151453.).

The weak correlation between TC activity in the entire NATL basin and the CAR index likely reflects the definition of this index, which is calculated solely using SST anomalies in the Caribbean Sea. Furthermore, the TSA index shows no direct relationship with cyclonic activity in the NATL, yet Pazos and Gimeno (2017)Pazos, M. & Gimeno, L. 2017. “Identification of moisture sources in the Atlantic Ocean for cyclogenesis processes”. In: 1st International Electronic Conference on Hydrological Cycle (ChyCle-2017). Sciforum Electronic Conference Series, Vol. 1, Basel, Switzerland: MDPI, DOI: 10.3390/CHyCle-2017-04862 and Pérez-Alarcón et al. (2020)Pérez-Alarcón, A.; Sorí, R.; Fernández-Alvarez, J. C.; Nieto, R. & Gimeno, L. 2020 . “Moisture Sources for Tropical Cyclones Genesis in the Coast of West Africa through a Lagrangian Approach”. Environmental Sciences Proceedings, 4:3, ISSN: 2673-4931, DOI: 10.3390/ecas2020-08126 demonstrated that the South Atlantic is an important source of moisture for TC genesis near the coast of West Africa. This behaviour supports an indirect influence of South Atlantic SST on cyclonic activity in NATL due to its effect on evaporation over the sea surface, which injecting moisture into the atmosphere.