No known hurricane has ever crossed the equator. Hurricanes require the Coriolis force to develop and generally form at least 5° away from the equator since the Coriolis force is zero there.

Are the physics of the earth and tropical systems such that it is impossible for a hurricane to cross the equator after forming, or are the forces working against this occurring so strong that an equator crossing hurricane is an exceedingly rare event we may not witness in 1000+ years?

Tracks and Intensity of All Tropical Storms

  • 10
    $\begingroup$ From what I understand of the matter it is physically impossible for a hurricane formed on one hemisphere to continue on the other hemisphere as the Coriolis effect (not a force!) is directed in the opposite direction. By crossing the equator the hurricane would stop turning, and only if the conditions are right on the other side of the equator (or should I say Intertropical Convergence Zone?) a new hurricane could form from the released energy of the original hurricane. But as this is mostly (an educated?) guess I don't think it deserves to be an answer. $\endgroup$
    – hugovdberg
    Commented Apr 17, 2014 at 7:37
  • 4
    $\begingroup$ Is it technically incorrect to say 'Coriolis Force' if it is only an apparent force, not a real one? I've generally known the effect to be referred to as a 'force' even if this is a misnomer. $\endgroup$
    – DrewP84
    Commented Apr 17, 2014 at 14:22
  • 4
    $\begingroup$ @DrewP84 its a force in our reference frame, and represents a term in our equations of motion in that reference frame so I think of it as correct. We also say "force" when we are really talking about accelerations (as we divide mass out of everything) so it can't be any worse than that. $\endgroup$
    – casey
    Commented Apr 17, 2014 at 14:37
  • 5
    $\begingroup$ Nice writeup on the subject at the Weather Underground. $\endgroup$
    – DrewP84
    Commented Apr 17, 2014 at 15:42
  • 10
    $\begingroup$ "The Coriolis effect is not a force" is a pretty similar argument to "there's no such thing as centrifugal force"... And that makes it nearly obligatory for me to link to xkcd.com/123 :-) $\endgroup$ Commented Apr 17, 2014 at 22:27

3 Answers 3



It is well known that the Coriolis force is needed to form a hurricane, and the figure of 5oN/S as the minimum for formation is widely publicized. You can also find record of tropical storm formation near India as far south as 1.4oN.

The problem of crossing the Equator isn't one of hurricane formation though, it is one of hurricane motion. Due to Coriolis, a hurricane initially moving parallel to the Equator will start gaining a poleward component to its motion, thus moving it away from the Equator. But, because this is due to Coriolis, if you could get a storm close enough to the Equator, this effect would not be as strong. This would be an improbable track, but I'm not willing to call it impossible. We haven't had satellites all that long, and all we can really say is that it hasn't happened since we've been watching.

If a storm did cross the equator though, what would it do? Nothing at first, but as it moved further into the opposite hemisphere, Coriolis would be working against the storm and it would spin down, become disorganized and cease to be a hurricane, probably becoming a remnant low.

A tropical disturbance has crossed the equator. One such disturbance occurred June 27, 2008 in the Atlantic basin (south to north) that retained its clockwise motion for some time:

animated satellite loop

  • 2
    $\begingroup$ This reflects my understanding as well. I'm curious about the evolution of the tropical system that formed near 1.4°N. Do you happen to know the year this occurred? There is a trajectory on the image above of a cyclone south of India that initially moves towards the equator, but quickly turns towards the northwest. Does the beta-effect increase in a hurricane the closer it gets to the equator, essentially making it more and more difficult to reach the equator? $\endgroup$
    – DrewP84
    Commented Apr 17, 2014 at 15:26
  • 4
    $\begingroup$ @DrewP84 That storm was Tropical Storm Vamei in 2001. There is an incomplete list of other storms at en.wikipedia.org/wiki/List_of_Equatorial_tropical_cyclones. My understanding of beta effect and hurricanes makes is that it will be reduced toward the equator. I'm also adding in an edit above a disturbance that did cross the equator. $\endgroup$
    – casey
    Commented Apr 17, 2014 at 15:52
  • $\begingroup$ Thanks! I noticed that animated loop from June 26-27, 2008. Very interesting! I read some speculation that it could have been a smaller mesolow feature rather than a true cyclone precursor. $\endgroup$
    – DrewP84
    Commented Apr 17, 2014 at 18:17
  • 1
    $\begingroup$ @DrewP84 I didn't mean to imply that specific disturbance developed into something more, just that in general a disturbance precedes a cyclone. I found some anecdotal mentions of another similar disturbance that did later develop into a cyclone, but I could not find any reference that I could actually follow up on, and the noise in googling for storms crossing the equator is enormous (as you might already have noticed). $\endgroup$
    – casey
    Commented Apr 17, 2014 at 18:22
  • $\begingroup$ @casey - looks like TC en.wikipedia.org/wiki/Cyclone_Agni has crossed the equator. $\endgroup$
    – user1066
    Commented Sep 7, 2015 at 3:01

The Coriolis effect is necessary for FORMATION but NOT for MAINTENANCE of a tropical cyclone.

Once formed, in a full-fledged tropical cyclone of hurricane intensity the wind balance is cyclostrophic, between the pressure-gradient and centrifugal force, with the Coriolis effect negligible by comparison.

This is especially true if the tropical cyclone is small but intense. Thus a full-fledged tropical cyclone of hurricane intensity would encounter NO difficulty if it crossed the equator.

Its circulation would be classified as baric before crossing the equator and antibaric afterwards. Thus a full-fledged tropical cyclone of hurricane intensity will NOT weaken upon crossing the equator due to the reversed Coriolis effect (although it may weaken or strengthen due to other causes).

An antibaric cyclone is fully consistent with the laws of motion. Except on the smallest scales such as dust devils or whirlpools it is difficult to initiate, but once initiated it can be stable.

A tropical cyclone crossing the equator may be the only way that a synoptic-scale antibaric cyclone can be initiated. If a tropical cyclone, especially a large one, crosses the equator and attains appreciable latitude in the opposite hemisphere, say the outer tropics or lower middle latitudes, then the gradient wind approximation rather than the cyclostrophic wind approximation is applicable.

But antibaric gradient wind is still a stable flow. See for example James R. Holton and Gregory J. Hakim, Dynamic Meteorology Fifth Edition Section 3.2.5 on pp. 74-77.

In baric gradient flow the Coriolis and centrifugal forces acting outwards balance the pressure-gradient force acting inwards.

In antibaric flow the centrifugal force acting outwards balances the pressure gradient and Coriolis forces acting inwards.

Thus for a given pressure gradient and radius of curvature of the isobars, the centrifugal force must be stronger, and hence the wind speed higher, for an antibaric tropical cyclone than for a baric one.

This does not mean that a tropical cyclone must intensify if it crosses the equator. Whether baric or anitbaric, its intensity is determined by the available energy or exergy corresponding to the sea-surface and tropopause temperatures (with possible deductions for wind shear, entrainment of dry air, land interaction, etc.)

The First and Second Laws of Thermodynamics require that the pressure gradient and radius of curvature of the isobars must adjust to the exergy supply --- not vice versa.

According to Holton and Hakim, an antibaric (synoptic-scale) hurricane can easily exist if it is centered say only a few degrees from the Equator in the opposite hemisphere, but this becomes more and more difficult and hence more and more unlikely with increasing latitude in the opposite hemisphere.

Whether at some latitude in the opposite hemisphere it becomes absolutely impossible may be an open question.

Small-scale antibaric cyclones (dust devils, whirlpools, waterspouts, and to a lesser extent tornadoes) can easily exist at any any latitude where sufficient exergy exists to generate and maintain them.

  • $\begingroup$ You got me there, once-upon-a-time I did quite a bit of TC research, including vortex profiles, and guess that knowledge faded away. That said, based upon this (not a true research reference, but pretty sharp), you're right that there is a cyclostrophic region... though it's only the core that is cyclostrophic (such that there's still trouble to the TC as it crossed)? You sound pretty knowledgeable of TC dynamics, can you edit to include a better reference for folks? But good information/answer, well done. $\endgroup$ Commented Jun 19, 2017 at 13:45
  • $\begingroup$ To your new add (about antibaric balance), as you say, v doesn't respond to create the balance... the balance has to be stable to subsist. And so as Coriolis diminishes when crossing, the flow will just become inward and the storm will weaken... and then as Coriolis increases in the opposite direction, there's still nothing at all to foster an increase in velocity to match. So it should just stray further and further from any balance (antibaric or otherwise). It's a fun idea, but it's not like a supercell that generates the vorticity actively to make it happen. $\endgroup$ Commented Oct 19, 2017 at 4:08
  • $\begingroup$ While to the best of my knowledge no tropical cyclone has ever crossed the equator, there have been some centered as close as about 1 degree of latitude from the equator. Thus almost half of the circulation was on the opposite side of the equator and hence antibaric, and operated just fine. We need one tropical cyclone to cross the equator to at least the outer tropics in the opposite hemisphere to provide an experimental test of full antibaricity. $\endgroup$
    – Jack Denur
    Commented Oct 19, 2017 at 7:19
  • $\begingroup$ I guess technically one inch across the equator the circulation would be antibaric... but in scale analysis, Coriolis would be negligible. Within a few degrees of the equator, it would basically be straightforward pgf/centrifugal balance, right? As I understand it, a TC ramps up as pressure falls due to latent heat release... then the isoallobaric wind flows inward, but Coriolis balances a percentage of it, leading to a portion becoming increased rotation. But once across the equator, Coriolis would oppose the existing rotation, so there'd be no way for v to increase... $\endgroup$ Commented Oct 19, 2017 at 16:03
  • 1
    $\begingroup$ An antibaric whirlpool can certainly exist if the reversed rotation of the carousel is sufficiently slow. As the speed of reversed rotation of the carousel is increased, it likely becomes more and more difficult. At what speed of reversed rotation, if any, it becomes absolutely impossible may be an open question. $\endgroup$
    – Jack Denur
    Commented Oct 22, 2017 at 23:05

Here is an archived comment on NASA’s Earth Observatory website, regarding an estimated once in 100-400 years equatorial cyclone – typhoon Varmei in December 2001.


  • 1
    $\begingroup$ Interesting, but nowhere does the article mention that the typhoon crossed the equator $\endgroup$
    – Fred
    Commented Aug 8, 2019 at 9:27
  • $\begingroup$ Indeed. I did come across a brief, non-technical comment by Gary Barnes, Professor of Meteorology at the University of Hawaii which appears to be in line with the general opinion in this forum that yes, a rotating storm could cross the equator, but there appears to be no record of one actually doing so. soest.hawaii.edu/GG/ASK/hurricanes.html $\endgroup$ Commented Aug 8, 2019 at 10:08
  • 1
    $\begingroup$ Your answer is in another castle. Please edit the essentials into your text. $\endgroup$
    – Jan Doggen
    Commented Aug 8, 2019 at 11:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.