Tropical cyclone frequency
Why does Earth have as many tropical cyclones as it does?
Summary
Why it is open
A tropical cyclone is a low-pressure system that forms over tropical waters, with thunderstorms near the centre of its closed, rotating winds [15]. In records supplied by the regional forecasting centres, an average of 84.3 storms a year reached tropical-storm strength (sustained winds above 17 m/s) from 1980 to 2024, with as few as 62 and as many as 105 in a single year [15].
In a 2021 review, Adam Sobel and colleagues wrote that "there is no accepted theory that explains the average number of TCs that occur each year on the Earth, nor how that number will change with global warming" [11]. Arguments from global budgets of heat or moisture do not yet appear helpful, they add, and nor does a detailed understanding of how individual storms form [11]. Kimberly Hoogewind and colleagues likewise describe the processes that control the number as unknown [7].
The problem is partly a dispute about what limits genesis, the formation of a new storm. One view holds that the large-scale environment decides; another adds that the supply of precursor disturbances, called "seeds", matters independently [11, 13]. Kerry Emanuel argued in 2022 that splitting genesis into seeds and a transition probability does not resolve this "long-standing issue" [13].
No body sets a bar for a resolution. The sources point to what one would need: a theory that predicts the absolute number, not only its variations in space and time [11]; an answer to whether the environment or the background weather controls the number [13]; and projections that agree on the sign of the change, which they do not [9, 11].
What is known
The observed record. Global records of tropical cyclones before geostationary weather satellites in the 1970s are more prone to discontinuities and sampling problems [14]. Of the few decades of reliable data, Savin Chand and colleagues wrote in 2022 that "there is no clear evidence of an observed trend in global TC numbers" [14].
Genesis indices (2007–2020). Empirical indices predict genesis from large-scale conditions. The genesis potential index of Emanuel and Nolan combines four of them: spin in the lower atmosphere (vorticity at 850 hPa), relative humidity at 600 hPa, vertical wind shear between 850 and 200 hPa, and potential intensity [1]. Suzana Camargo, Emanuel and Sobel found that the index, composited over El Niño and La Niña years, qualitatively replicates the observed year-to-year variations in genesis frequency and location in several basins [1]; see the Papex problem Dynamics of the El Niño–Southern Oscillation. Such indices explain some of the relative variations in frequency, "but not its absolute value" [11]. Hoogewind and colleagues filled the regions and seasons whose thermodynamic conditions favour tropical cyclones with as many storms as fit at a fixed spacing; this "maximum potential genesis" rate was an order of magnitude larger than the observed rate, and its year-to-year changes correlate poorly with the observed global count [7].
Seeds (2018–2020). Gabriel Vecchi and colleagues found that in their models global frequency responds both to the number of seeds, which rises with warming and falls with higher carbon dioxide, and to how efficiently seeds develop into storms [5]. Tsung-Lin Hsieh and colleagues built a framework in which rotating seeds arise from non-rotating convective clusters, whose number follows the large-scale vertical motion, with a transition probability set by large-scale vorticity; across a hierarchy of models, the response of tropical cyclones to perturbed sea surface temperature and carbon dioxide followed the response of seeds [8]. In regional simulations by Christina Patricola and colleagues, removing the African easterly waves, from which Atlantic storms readily originate, produced no significant change in the seasonal number of Atlantic storms [4].
Idealised worlds (2013–2020). Marat Khairoutdinov and Emanuel simulated radiative–convective equilibrium, an idealisation of the mean tropical climate, with constant rotation and sea surface temperatures from 21 to 36 °C; their "TC World" held from 8 to 26 storms, fewer at higher temperatures, and the spacing between storm centres appeared to scale with the deformation radius, a natural length scale of a rotating atmosphere [3]. In an aquaplanet, a model Earth covered by ocean, Timothy Merlis, Ming Zhao and Isaac Held found that hurricane frequency rose with warming and argued that this was because the tropical rain belt moved poleward; with the rain belt held in place, frequency fell [2]. Hsieh and colleagues tested parts of their framework in aquaplanets with shifted rain belts and with uniform sea surface temperature [8].
Projections (2019–2022). In a 2020 assessment, Thomas Knutson and colleagues placed a decrease in global frequency, "as projected in most studies", among the projections on which author opinion was more mixed and confidence lower [9]. By 2021, global models with grid spacings of 25–50 km had much improved the simulation of tropical cyclones; many projected fewer storms with warming, some more [11]. In 2019, in three coupled models with increasingly fine grids, doubled carbon dioxide reduced global frequency substantially at about 50 km resolution and changed it insignificantly at about 25 km [5]. In 2021 Emanuel downscaled the sixth generation of coupled climate models (CMIP6) with a very high-resolution but simplified tropical cyclone model and found increases in frequency that were robust across the models [10]. Where storms occur may also change: reviewing storm physics against the responses of the Hadley circulation, jet streams and tropical rain belt to warming, Joshua Studholme and colleagues concluded that twenty-first-century storms will most probably span a broader range of latitudes, as conditions in mid-latitudes become more favourable, although precise estimates remain beyond current methods [12]; see Dynamics of storm tracks.
Recent claims. In 2022 Chand and colleagues reconstructed annual counts back to 1850 from the Twentieth Century Reanalysis and reported that the global number fell by about 13% in the twentieth century relative to 1850–1900, and a larger decline, about 23%, evident after about 1950, consistent with a weakening of the Hadley and Walker circulations [14]. They noted that the number of surface observations assimilated into the reanalysis increased after about 1946, a concern for homogeneity, and that a second reanalysis also showed the decline [14]. They also pointed to a downscaling study that had reported an increasing North Atlantic trend over the past century, and argued that such methods assume, implicitly or explicitly, that adequate seeds are always available, that their rate is independent of climate, or that it is controlled by the same factors as their development into storms [14].
Main approaches
Environmental control. The regional environment mostly controls tropical cyclone climatology, as Emanuel argues [13], and genesis indices and downscaling models are built on it [1, 10]. Strongest evidence: indices replicate the effects of El Niño and La Niña on genesis [1]; without African easterly waves, the Atlantic produced no significantly different number of storms [4]. Main difficulty: indices do not give the absolute number [11], and in one downscaling system the choice of humidity variable reversed the sign of the projected change, an uncertainty its authors call "fundamental and epistemic in nature" [6].
Seed control. The genesis rate is the frequency of seeds times the probability that a seed becomes a storm [8, 13]. Strongest evidence: across a hierarchy of models, storms followed seeds when sea surface temperature and carbon dioxide were perturbed [8]. Main difficulty: the consistency broke down when sea surface temperature gradients were strongly enhanced or reduced [8], and the split between seeds and probability depends on how seeds are defined [13].
Explicit simulation. Global models with fine enough grids simulate the storms themselves [11]. Strongest evidence: grid spacings of 25–50 km have much improved the simulation of tropical cyclone activity [11]. Main difficulty: the models disagree on the sign of the change [11], and the response depends on resolution and on each model's sea surface temperature biases [5].
Idealised hierarchies. Simplified worlds let one factor be varied at a time [11]. Strongest evidence: the counts respond in clear ways to a single change: sea surface temperature in TC World [3], the position of the rain belt in the aquaplanet [2]. Main difficulty: TC World needed rotation about ten times stronger than in the tropics to fit several storms into its domain [3].
Open sub-questions
- The absolute number. Can an index or theory reproduce the observed average of 84.3 storms a year without being fitted to counts [11, 15]?
- Which humidity. In experiments where column relative humidity and saturation deficit diverge, which better predicts genesis, and which sign of change does it give [6]?
- Seed definitions. Does the seed-based explanation of the response to warming hold under the alternative seed definitions Emanuel examined [8, 13]?
- Seeds elsewhere. Would suppressing the precursor disturbances of another basin also leave its seasonal storm numbers unchanged, as for African easterly waves in the Atlantic [4]?
- Spacing and the ceiling. Does storm spacing set by the deformation radius [3] explain part of the gap between maximum potential genesis and the observed rate [7]?
- The century-scale decline. Does the twentieth-century decline survive reconstructions that are insensitive to the post-1946 rise in assimilated observations [14]?
Work on this in Papex
If you have a testable answer to one of the sub-questions above, you can draft it as a private hypothesis in Papex, or prepare the draft from Claude or ChatGPT with the Papex connector. You review the exact version, sign it and decide whether to publish it; a public record that names this problem is listed on this page. A Papex record shows who signed which version and when. It is not peer review and does not mean the hypothesis is correct.
Key references
- Camargo, 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(19), 4819–4834. 10.1175/JCLI4282.1 · Checked 16 Oct 2026
- Merlis, T. M., Zhao, M. & Held, I. M. (2013). The sensitivity of hurricane frequency to ITCZ changes and radiatively forced warming in aquaplanet simulations. Geophysical Research Letters 40(15), 4109–4114. 10.1002/grl.50680 · Checked 16 Oct 2026
- Khairoutdinov, M. & Emanuel, K. (2013). Rotating radiative-convective equilibrium simulated by a cloud-resolving model. Journal of Advances in Modeling Earth Systems 5(4), 816–825. 10.1002/2013MS000253 · Checked 16 Oct 2026
- Patricola, C. M., Saravanan, R. & Chang, P. (2018). The response of Atlantic tropical cyclones to suppression of African easterly waves. Geophysical Research Letters 45(1), 471–479. 10.1002/2017GL076081 · Checked 16 Oct 2026
- Vecchi, G. A., Delworth, T. L., Murakami, H., Underwood, S. D., Wittenberg, A. T., Zeng, F. et al. (2019). Tropical cyclone sensitivities to CO2 doubling: roles of atmospheric resolution, synoptic variability and background climate changes. Climate Dynamics 53(9–10), 5999–6033. 10.1007/s00382-019-04913-y · Checked 16 Oct 2026
- Lee, C.-Y., Camargo, S. J., Sobel, A. H. & Tippett, M. K. (2020). Statistical–dynamical downscaling projections of tropical cyclone activity in a warming climate: two diverging genesis scenarios. Journal of Climate 33(11), 4815–4834. 10.1175/JCLI-D-19-0452.1 · Checked 16 Oct 2026
- Hoogewind, K. A., Chavas, D. R., Schenkel, B. A. & O'Neill, M. E. (2020). Exploring controls on tropical cyclone count through the geography of environmental favorability. Journal of Climate 33(5), 1725–1745. 10.1175/JCLI-D-18-0862.1 · Checked 16 Oct 2026
- Hsieh, T.-L., Vecchi, G. A., Yang, W., Held, I. M. & Garner, S. T. (2020). Large-scale control on the frequency of tropical cyclones and seeds: a consistent relationship across a hierarchy of global atmospheric models. Climate Dynamics 55(11–12), 3177–3196. 10.1007/s00382-020-05446-5 · Checked 16 Oct 2026
- Knutson, T., Camargo, S. J., Chan, J. C. L., Emanuel, K., Ho, C.-H., Kossin, J. et al. (2020). Tropical cyclones and climate change assessment: Part II: projected response to anthropogenic warming. Bulletin of the American Meteorological Society 101(3), E303–E322. 10.1175/BAMS-D-18-0194.1 · Checked 16 Oct 2026
- Emanuel, K. (2021). Response of global tropical cyclone activity to increasing CO2: results from downscaling CMIP6 models. Journal of Climate 34(1), 57–70. 10.1175/JCLI-D-20-0367.1 · Checked 16 Oct 2026
- Sobel, A. H., Wing, A. A., Camargo, S. J., Patricola-DiRosario, C. M., Vecchi, G. A., Lee, C.-Y. & Tippett, M. K. (2021). Tropical cyclone frequency. Earth's Future 9(12), e2021EF002275. 10.1029/2021EF002275 · Checked 16 Oct 2026
- Studholme, J., Fedorov, A. V., Gulev, S. K., Emanuel, K. & Hodges, K. (2022). Poleward expansion of tropical cyclone latitudes in warming climates. Nature Geoscience 15(1), 14–28 (published online 2021). 10.1038/s41561-021-00859-1 · Checked 16 Oct 2026
- Emanuel, K. (2022). Tropical cyclone seeds, transition probabilities, and genesis. Journal of Climate 35(11), 3557–3566. 10.1175/JCLI-D-21-0922.1 · Checked 16 Oct 2026
- Chand, S. S., Walsh, K. J. E., Camargo, S. J., Kossin, J. P., Tory, K. J., Wehner, M. F. et al. (2022). Declining tropical cyclone frequency under global warming. Nature Climate Change 12(7), 655–661. 10.1038/s41558-022-01388-4 · Checked 16 Oct 2026
- NOAA Atlantic Oceanographic and Meteorological Laboratory, Hurricane Research Division. Hurricane FAQ: "What Is a Tropical Cyclone, Tropical Disturbance, Tropical Depression, Tropical Storm, Hurricane, and Typhoon?" and "What Are the Record Number of Tropical Storms in a Given Year by Basin?" (data for 1980–2024). Source · Checked 16 Oct 2026
Source
Open, as listed by Wikipedia · Unsolved problems in geoscience. Checked 29 Sep 2026.
The status is the source’s. Papex does not decide whether a problem is solved.
Records on this problem
No public record addresses this problem yet.