Why do hurricanes always turn north, away from the Equator?

[physicsponderer]

Hurricanes tend not to move towards the Equator for very long, turning north fairly soon, and have never been known to reach it, I think. Likewise cyclones in the southern hemisphere, which turn south before reaching it. Why is this? I have two tentative explanations.

1. Hurricanes start out moving west, or south west, carried by the Trades (Trade Winds) and the Coriolis Effect makes them veer right, which is north.

2. The air being sucked in from the south hits the hurricane more directly than the air being sucked in from the north, because the Coriolis effect is stronger the further from the Equator you get. The air from the north strikes a glancing blow, and adds to the rotation more, but pushes the hurricane less, in other words it contributes more to the angular momentum and less to the (linear) momentum of the hurricane, compared to the air coming from the south. Of course, "sucked in" is short for "pushed in by the air at normal atmospheric pressure".

 

[DaveC426913]

Why is this? I have two tentative explanations.

What you tried Googling 'what causes hurricanes to veer'?

 

[physicsponderer]

What you tried Googling 'what causes hurricanes to veer'?

I read about it on the internet. It was confusing, and the Coriolis effect was mentioned, so it might have inspired my first guess. The second guess is my own idea.

 

[DaveC426913]

The second guess is my own idea.

Which also leverages the Coriolis Force.

 

[anorlunda]

Not an answer to your question, but I just found this very beautiful graphic.
https://geozoneblog.files.wordpress.com/2014/05/globalwrappedclipped.png

Note that there appears to be a single outlier that hit Brazil.

 

[physicsponderer]

Which also leverages the Coriolis Force.

True. Silly me. On the other hand I am pretty sure the second is idea is new, apart from involving the Coriolis force.

 

[physicsponderer]

Not an answer to your question, but I just found this very beautiful graphic.
https://geozoneblog.files.wordpress.com/2014/05/globalwrappedclipped.png

View attachment 274625

Note that there appears to be a single outlier that hit Brazil.

Yes, that is a nice bit of eye candy.The aversion to the Equator is very clear, too.

 

[physicsponderer]

I just noticed that that remarkably straight line in your image is approximately level with the southern limit of the Sahara desert, which is also remarkably straight. I wonder whether there's a connection?

 

[DaveC426913]

True. Silly me. On the other hand I am pretty sure the second is idea is new, apart from involving the Coriolis force.

I am not 100% sure of this, but I'm mostly sure that the force that causes a hurricane to rotate and to veer away from the equator will operate on any material (in principle), not just air masses. Like, say, if you had a 1000km long pool noodle. i.e. it doesn't depend on winds or air pressure gradients.

 

[physicsponderer]

Correction: When I say I am pretty sure the second idea is new, I mean I'm pretty sure I came up with the idea on my own, not that no one else ever said it.

 

[physicsponderer]

I am not 100% sure of this, but I'm mostly sure that the force that causes a hurricane to rotate and to veer away from the equator will operate on any material (in principle), not just air masses. Like, say, if you had a 1000km long pool noodle. i.e. it doesn't depend on winds or air pressure gradients.

So how exactly does it work?

 

[PeterDonis]

Hurricanes start out moving west, or south west, carried by the Trades (Trade Winds) and the Coriolis Effect makes them veer right, which is north.

This will be a valid effect, yes. It would apply to any westward-moving object whose motion was not constrained.

Note, however, that this northward movement is actually just one component of the effect. The acceleration vector is actually "inward", towards the axis of rotation. On the equator, this direction would be straight down (which would basically amount to no effect for something like a hurricane that can't "fall down"--it can, however, affect the pressure gradient with altitude). North of the equator, this vector acquires a northward component that increases as you get further from the equator.

The air being sucked in from the south hits the hurricane more directly than the air being sucked in from the north, because the Coriolis effect is stronger the further from the Equator you get.

You are ignoring the fact that the hurricane itself rotates. A hurricane is a low pressure area, and in the northern hemisphere, circulation around a low pressure area is counter-clockwise. So air is not being "sucked in" from any direction in a straight line to begin with.

You might want to think about why the circulation around a low pressure area in the northern hemisphere is counter-clockwise, whereas around a high pressure area in the northern hemisphere it is clockwise. Note that one of these (which one?) is the opposite from the sense of rotation that consideration of the Coriolis effect in isolation would lead you to expect.

 

[DaveC426913]

Here is an article that attempts to explain it, but huge caveat: I cannot attest to its accuracy.
https://www.sjsu.edu/faculty/watkins/hurricanepaths.htm

 

[PeterDonis]

huge caveat: I cannot attest to its accuracy.

The basic math works, in the sense that if you take a vertical gyroscope axis and an eastward torque applied to it (to keep it vertical as the Earth rotates eastward), the precession is northward.

However, a hurricane is not a gyroscope or even very close to one; it is not a solid object and internal forces between its parts, to the extent they exist at all, are not at all similar to those in solid objects. And those internal forces in a solid object are what drive the precession of a gyroscope. (Feynman had a good analysis of this which I think appears in his lectures on physics.) In short, "gyroscope" is not a magic word that applies to anything with a rotation axis; it implies a particular model of the object which is not a very good model of a hurricane.

 

[PeterDonis]

(Feynman had a good analysis of this which I think appears in his lectures on physics.)

Here it is:

https://www.feynmanlectures.caltech.edu/I_20.html

 

[DaveC426913]

The basic math works, in the sense that if you take a vertical gyroscope axis and an eastward torque applied to it (to keep it vertical as the Earth rotates eastward), the precession is northward.

However, a hurricane is not a gyroscope or even very close to one; it is not a solid object and internal forces between its parts, to the extent they exist at all, are not at all similar to those in solid objects. And those internal forces in a solid object are what drive the precession of a gyroscope. (Feynman had a good analysis of this which I think appears in his lectures on physics.) In short, "gyroscope" is not a magic word that applies to anything with a rotation axis; it implies a particular model of the object which is not a very good model of a hurricane.

Cool. So does that put the lie to my Post #9 (with the pool noodle), too?

 

[PeterDonis]

does that put the lie to my Post #9 (with the pool noodle), too?

Not necessarily. Your general idea that whatever acts on hurricanes will act on other objects as well is correct. And, as I noted in post #12, the first idea in the OP, that a westward-moving object in the northern hemisphere will acquire a northward velocity due to the Coriolis effect, is correct. This works just as well for the average motion of an object made up of lots of little parcels with weak internal forces between them.

The real limitation of all these speculations is that they fail to recognize other effects on hurricanes, such as what other air masses are present, what the prevailing winds are at various altitudes, etc., all of which might well be stronger than the Corolis effect. The fact that these other effects are significant should be obvious from looking at the actual tracks of individual hurricanes; many of them do not follow the "start moving west, curve north, veer back east as they go north" pattern that the basic analysis we are discussing implies.

 

[Vanadium 50]

Note that there appears to be a single outlier that hit Brazil.

Also, lore is that cyclonic storms don't form near the equator (where the Coriolis force is zero), but every so often there is one, like Typhon Vamei in 2001.

 

[Borek]

I just noticed that that remarkably straight line in your image is approximately level with the southern limit of the Sahara desert, which is also remarkably straight. I wonder whether there's a connection?

https://en.wikipedia.org/wiki/Hadley_cell

 

[physicsponderer]

https://en.wikipedia.org/wiki/Hadley_cell

I was wondering about that, too, because the Hadley cell is said to cause the Sahara desert (but it didn't say how, nor why the Sahara doesn't extend all the way around the world, like the Hadley cell). But the southern edge of the Sahara is not at either end of the Hadley cell, but rather about one third up it which is not significant line in it, as far as I know. Also, I do wonder whether something as vague as the Hadley cell (correct me if I'm wrong) which is, supposedly, often "displaced" could cause that ruler-straight line in the Atlantic that is the southern limit of the zone visited by hurricanes.

 

[physicsponderer]

I wonder whether a hot air balloon could drift from a hurricane zone to the Equator? What about smoke? If so, the question is, why those things and not hurricanes? Also, what about a cloud? A storm? A former hurricane that is too weak to count as a hurricane anymore? Has any of these ever reached the Equator?

 

[physicsponderer]

What is meant by reaching the Equator? How to define that? If I have one hand over the Equator, I have surely reached it, but I suspect that with a hurricane the centre (meaning the centre of the eye?) must reach it.

 

[physicsponderer]

Also, lore is that cyclonic storms don't form near the equator (where the Coriolis force is zero), but every so often there is one, like Typhon Vamei in 2001.

It seems you are right. It was much nearer than I thought possible, only about 200 km from the Equator. And if it was (hypothetically) 500 km in radius, then in a sense it touched or even crossed it, perhaps. But the diagram showed only the path of its centre or centre of mass or something.

 

[physicsponderer]

You are ignoring the fact that the hurricane itself rotates. A hurricane is a low pressure area, and in the northern hemisphere, circulation around a low pressure area is counter-clockwise. So air is not being "sucked in" from any direction in a straight line to begin with.

You might want to think about why the circulation around a low pressure area in the northern hemisphere is counter-clockwise, whereas around a high pressure area in the northern hemisphere it is clockwise. Note that one of these (which one?) is the opposite from the sense of rotation that consideration of the Coriolis effect in isolation would lead you to expect.

I didn't forget about the rotation of the hurricane. Surely the air far away from the hurricane is essentially stationary to start with, and then it feel the tug from the hurricane and moves straight towards it at first?

Are you referring to the way one might wrongly expect Coriolis-caused bending to the right to cause clockwise rotation of the winds? That is a bit tricky to understand. I think I mastered it a while back though.

 

[PeterDonis]

Surely the air far away from the hurricane is essentially stationary to start with

No, it isn't, because there are other weather systems and effects besides the hurricane that are present. You yourself mentioned the trade winds. There is also the jet stream, plus other high or low pressure areas that might exist.

then it feel the tug from the hurricane

There is no "tug from the hurricane" that just turns on at some specified distance and is zero outside it. There is airflow driven by pressure differences everywhere that there are pressure differences, which is pretty much everywhere.

 

[physicsponderer]

Has a hurricane ever been observed going south east near the Equator? If it bent to the right, it would reach the Equator. So what would it do?

 

[PeterDonis]

If it bent to the right, it would reach the Equator. So what would it do?

Um, keep going?

You don't seem to have grasped previous statements that the Coriolis force is not the only thing that determines how hurricanes move.

 

[Vanadium 50]

You seem to think a Hurricane is a point object. They are big. When they (rarely) form near the equator, they influence the weather on both sides. You can't just talk about the eye, because many storms don't have a well-formed or visible eye.

Here's Wikipedia's picture of Cyclone Agni. Can you spot the eye? And isn'tg it pretty clearly in both hemispheres?

 

[physicsponderer]

Here is an article that attempts to explain it, but huge caveat: I cannot attest to its accuracy.
https://www.sjsu.edu/faculty/watkins/hurricanepaths.htm

This is very interesting. It seems to be very authoritative. I am struggling with the jargon though. In particular, I couldn't find anywhere a definition of 'global torque'. The fact that there is a company called 'Global Torque' does not help. Here's the abstract:
Abstract: When an entity with angular momentum about its axis, such as a gyroscope, is subjected to a torque it precesses. If it is forced to precess it experiences a torque. Tropical cyclones (hurricanes, typhoons etc.) have angular momenta with respect to their axes. They turn with the Earth's rotation while the verticality of their angular momentum is enforced by the rising of the warm, moist air at their centers. Their angular momentum vectors are thus forced to precess. This results in their being subject to a global torque which accelerates them toward the Earth's pole in their hemisphere. Hurricanes are thus accelerated toward the North Pole.
There appears to be no other explanation of why tropical storms move poleward.
These tropical cyclones also have angular momentum with respect to the Earth's axis. The conservation of this angular momentum means that as they move to higher latitudes they experience an acceleration to the east. Thus a hurricane that develops in the low latitude Atlantic and appears to move west begins to turn north. It thus recurves to the east as it moves north. A Southern Hemispheric tropical cyclone such as an Australian willy-willy moves west and then recurves to the southeast.

Thanks a lot for posting this.

 

[physicsponderer]

You seem to think a Hurricane is a point object. They are big. When they (rarely) form near the equator, they influence the weather on both sides. You can't just talk about the eye, because many storms don't have a well-formed or visible eye.

Here's Wikipedia's picture of Cyclone Agni. Can you spot the eye? And isn'tg it pretty clearly in both hemispheres?

View attachment 274678

'You seem to think a Hurricane is a point object. '
Have you read the whole of this thread?

 

[physicsponderer]

To Vanadium:

It seems you are right. It was much nearer than I thought possible, only about 200 km from the Equator. And if it was (hypothetically) 500 km in radius, then in a sense it touched or even crossed it, perhaps. But the diagram showed only the path of its centre or centre of mass or something.

I posted the above in reply to one of your posts, so obviously I know that hurricanes are big, and not points.

 

[physicsponderer]

You seem to think a Hurricane is a point object. They are big. When they (rarely) form near the equator, they influence the weather on both sides. You can't just talk about the eye, because many storms don't have a well-formed or visible eye.

Here's Wikipedia's picture of Cyclone Agni. Can you spot the eye? And isn'tg it pretty clearly in both hemispheres?

View attachment 274678

The centre would seem to be about two degrees north of the Equator. What is making the wind spiral inwards I wonder? That is a very interesting picture. I thought that wasn't supposed to be possible. I guess by chance there was enough anticlockwise angular momentum locally so the Coriolis effect was not needed.

 

[PeterDonis]

This is very interesting. It seems to be very authoritative.

It's not. It's someone's personal site, not a textbook or peer-reviewed paper.

As for the content, see my comments in post #14.

 

[PeterDonis]

obviously I know that hurricanes are big, and not points

Maybe you know it, but you don't seem to be thinking very carefully about its implications. To name just one: since the direction of airflow (wind) varies from one place in the hurricane to another, whatever Coriolis effect exists on the airflow will vary from one place in the hurricane to another as well. So considering the hurricane to be a single object with a single Coriolis effect applied to it is obviously wrong. Yet that is what you are trying to do.

 

[physicsponderer]

Maybe you know it, but you don't seem to be thinking very carefully about its implications. To name just one: since the direction of airflow (wind) varies from one place in the hurricane to another, whatever Coriolis effect exists on the airflow will vary from one place in the hurricane to another as well. So considering the hurricane to be a single object with a single Coriolis effect applied to it is obviously wrong. Yet that is what you are trying to do.

Where did I try to do that?