# Curious Question about Atmospheric Pressure.

Only_Curious
I am new here and I have a question about pressure. My idea about atmospheric pressure is that it is a pressure caused by the force directed downward due to the weight of the atmosphere. If that is the case, does it mean that we doesn't feel any pressure horizontally? Please correct me if I am wrong with how I understand this. Thanks!

The important thing is not to stop questioning. - Albert Einstein

Lsos
Imagine pushing vertically on an inflated balloon. You will see that it bulges out on all sides, and pretty much everywhere it can...not just vertically.

So yes, it is the weight of the atmosphere that causes pressure, but it doesn't just work vertically. It works in all directions, even up.

Only_Curious
Thanks for the reply Lsos. Actually I get your point, but the thing is that it only works for confined fluids as stated by Pascal's principle. So you are saying that we can think of our atmosphere as a confined fluid? Or is there other explanation on this?

Homework Helper
Actually I get your point, but the thing is that it only works for confined fluids as stated by Pascal's principle. So you are saying that we can think of our atmosphere as a confined fluid?
The atmoshpere is confined by the surface of the Earth and the force of gravity that pulls it towards the surface of the earth.

Only_Curious
Thanks for clearing up rcgldr

Gold Member
Any region of the atmosphere is 'confined' by the regions around it, which exert a pressure inwards on it - whether you talk of the ground, the air on top or columns of air around it. Because the Earth is a rough spheroid, you don't need 'walls' around any particular piece. This is analogous to the tension in the envelope of a balloon, which isn't attached'' to anything but itself. The balloon stays up as long as the envelope is intact.

klimatos
I am new here and I have a question about pressure. My idea about atmospheric pressure is that it is a pressure caused by the force directed downward due to the weight of the atmosphere. If that is the case, does it mean that we doesn't feel any pressure horizontally? Please correct me if I am wrong with how I understand this. Thanks!

The important thing is not to stop questioning. - Albert Einstein

Atmospheric pressure is the pressure exerted on any exposed surface by the impact of air molecules upon that surface. It is the simple product of the number of impacts per square meter per second and the mean impulse per impact. In still air, that pressure is essentially the same in all directions.

Under conditions of equilibrium, it may be shown to approximate the "weight" of the overlying air (the barometric equation).

When the wind is blowing, that approximation becomes increasingly less accurate.

Gold Member
If the wind is blowing then it must be from a higher pressure region to a lower pressure region. Somewhere, something must be forcing air into that high pressure region and something else must be causing the low pressure region. What you say (as in that other interminable thread) does not consider that. The fact that moving air involves a pressure difference is not relevant if the mean pressure over the region where all the 'weather' is occurring remains the same. Have you a reason why the mean pressure should not the same? You have never stated one, so far. I think you need to look further at the problem rather than just consider the Bernoulli effect in one place.
One may need to add the weight of all aircraft that happen to be in the air to the total weight of the atmosphere to get a more accurate value for mean pressure. (And all the budgies, too?)

klimatos
If the wind is blowing then it must be from a higher pressure region to a lower pressure region.

There are lots of winds that do not blow from areas of high pressure to areas of low pressure. Virtually all gravity winds blow from areas of low pressure to areas of high pressure. If you think of winds as three-dimensional phenomena (the only rational perspective, in my opinion) then all areas when the air is sinking represent movement from low pressure areas to high pressure areas.

Let us be careful here to talk about actual pressures (i. e., those measured by a manometer), not the fictional pressures that result when actual pressures are "reduced to sea level".

klimatos
The fact that moving air involves a pressure difference is not relevant if the mean pressure over the region where all the 'weather' is occurring remains the same. Have you a reason why the mean pressure should not the same? You have never stated one, so far.

You've lost me here completely. The OP wanted to know why atmospheric pressures were omnidirectional when the "weight of the atmosphere" is unidirectional (down). I think I answered that question.

How did we get from that to the "mean pressure over the region where all the weather is occurring" (whatever that is) or "why the mean pressure should not [be] the same?" What mean pressure are we talking about?

A barometer measures the mean air pressure on its sensing surface during the response time of the instrument. This sensing surface is quite small. I know of no instruments that measure pressure over "regions".

As for atmospheric pressures "staying the same" when weather is occurring, you must be dealing with a different atmosphere than I have been. It is my experience that atmospheric pressures vary widely--both spatially and temporally--when weather is occurring.

Gold Member
I really have no idea how a gas can move for any other reason than pressure difference. Could you explain, please? Can it be pulled or twisted? What is a "gravity wind"?
By mean pressure I mean the mean of all pressures over a large region. There will be regions of high pressure and regions of low pressure and there will be regions of intermediate pressure. This will vary for different altitudes as well. You seem to be pinning the whole of your argument on the Bernouli effect on a particular body of air which is moving. You seem to be ignoring any air movement above or any pressure differences elsewhere. That is not the whole picture. Behaviour in one specific region does not justify an overall theory. You need to justify it in the wider context.

You are right to point out the OP but it doesn't make your statements right.

For the OP:
Pressure in fluids acts equally in all directions over a small volume. The molecules are moving randomly throughout the volume and have equal motion in all directions. If you explain the pressure in terms of the average change of momentum (as the molecules bounce off the side of a notional box, containing the gas), it is the same in all directions. This applies when the body of gas is stationary. When the gas is moving in a given direction, the pressure may not be the same in all directions (e.g. the blast from a jet engine)

klimatos
I really have no idea how a gas can move for any other reason than pressure difference. Could you explain, please? Can it be pulled or twisted? What is a "gravity wind"?

There will be a net flow of air from A to B whenever the molecular flux from A to B is greater than the molecular flux from B to A. This is most obvious with pressure differences, but can occur under isobaric conditions as a result of density differences, temperature differences, and humidity differences.

Gravity winds (also known as katabatic winds) occur as a result of density differences. Glacier winds are prime examples of such winds, and can reach speeds in excess of 100 mph. In essence, cold air flows down slopes, even under isobaric conditions. Other examples of gravity winds include the chinook, the Santa Ana, the Mistral, the Bora, the Foehn, and many others. As a petty example, when you open the refrigerator door, cold air flows out onto your feet. It does this despite there being no significant difference in air pressure.

Under isobaric conditions, cold air will flow towards warm air. At 1000 hPa, the flux rate for air at -25°C is 3.11 x 10^27 molecules per square meter per second. At +25°C, the flux is 2.84 x 10^27. This flux differential will manifest itself in a flow of air from the colder to the warmer. This is the genesis of the very common daily alternations of "land breezes" and "sea breezes".

Finally, there will be a net flow of air away from evaporating surfaces and toward condensing surfaces.

These are all common illustrations of isobaric winds.

klimatos
Behaviour in one specific region does not justify an overall theory. You need to justify it in the wider context.

Which overall theory are you referring to? Kinetic gas theory? Statistical mechanics? If you will specify my overall theory, I shall do my best to justify it.

As I can recall, I made three statements in my original response to the OP. In the first, I used kinetic gas theory and statistical mechanics to explain why atmospheric pressures were omnidirectional. Surely you don't object to those well-accepted perspectives?

In my second, I stated that atmospheric pressures approximated the "weight of the overlying atmosphere". This is Atmospheric Science 101. Do you object to this?

In my third, I stated that that approximation was degraded by increased wind velocities. Is this the one that "sticks in your craw"? That is simple application of the Bernoulli Theorem.

Gold Member
@klimatos

Unless you are suggesting that some sort of anti gravity exists, you cannot produce less mean force on the surface of the Earth than that which corresponds to the weight of the fluid above it. For every area of low pressure there must be some high pressure somewhere else. Certainly this must be true for the long-term time average.

[response to spambot deleted]

Klimatos: Of course I was objecting to your use of Bernouli, in isolation and some other meteorological terms (as I already said). I already included simple kinetic theory on one of my posts so I am hardly likely to be questioning that.

A U tube containing liquids of different densities will only be balanced when the pressures at the bottom (or anywhere else, for that matter) are equal. What causes a flow is pressure, caused by density times column height or modified by mean velocity. It's forces (pressure differences) that make things move - not density or temperature on their own. You are using 'shorthand' terms, used by Meteorologists (and well understood by the best of them, I'm sure), to attempt a Physical explanation. It can't work that way round. In the end, there needs to be an explanation based entirely on Physics.

Last edited by a moderator:
Mentor
There are lots of winds that do not blow from areas of high pressure to areas of low pressure. Virtually all gravity winds blow from areas of low pressure to areas of high pressure.
No. Air cannot flow through an increasing pressure gradient. Gravity causes the pressure gradient that moves the air in the case of gravity winds: http://en.wikipedia.org/wiki/File:Katabatic-wind_hg.png

This should be obvious: An area of cold and dense air over a glacier is more dense and thus has more weight than the air next to glacier. So a higher pressure than air at the same elevation, next to it. A gravity wind is really just the reverse of a thermal.

You should replace the air in your refrigerator example with water, for higher contrast, and re-examine...

Last edited:
klimatos
No. Air cannot flow through an increasing pressure gradient. Gravity causes the pressure gradient that moves the air in the case of gravity winds: http://en.wikipedia.org/wiki/File:Katabatic-wind_hg.png

This should be obvious: An area of cold and dense air over a glacier is more dense and thus has more weight than the air next to glacier. So a higher pressure than air at the same elevation, next to it.

Okay, Russ. Go to your link. Measure the ambient pressure at the tail of the blue arrow (A). Measure the ambient pressure at the head of the arrow (B). You will find that the pressure at A is less than the pressure at B, but that the air flows from A to B.

Every time there is subsidence in the atmosphere (the downward leg of a Hadley Cell, for instance) you have air moving from an area of lower real pressures (not pressure reduced to sea level) to an area of higher real pressures. All that is required is a difference in density.

The same thing happens in water. When a mass of cold water sinks it is moving from an area of lower pressure to an area of higher pressure.

Mentor
You are allowing the fact that a vertical pressure gradient exists in the atmosphere to confuse the issue. The way you characterize it is imprecise and misleading at best. And looking back, you did something similar with a definition of "atmospheric pressure" earlier by making it too generic so that it implies velocity pressure is a component of it. And now that I think about it, I think we had this problem once before in another thread.

Last edited:
klimatos
@klimatos and rock100

1) Unless you are suggesting that some sort of anti gravity exists, you cannot produce less mean force on the surface of the Earth than that which corresponds to the weight of the fluid above it.

2) For every area of low pressure there must be some high pressure somewhere else. Certainly this must be true for the long-term time average.

1) Sure you can. All you have to do is put the fluid in motion. The atmosphere is almost always in motion.

2) I am more interested in explaining local weather phenomena than in global averages.

Addendum) I suspect that I am dealing with laboratory scientists here. The free atmosphere does not always behave as do samples of air in the laboratory. To me, atmospheric physics is primarily an observational science like astronomy, rather than a laboratory science like classical physics.

I don't understand why both you and russ have such a hard time in believing that cold air will flow down slopes, from measured lower pressure areas (high elevation) to measured higher pressure areas (lower elevation). Cold water flows down underwater slopes in exactly the same manner.

I'll bet you could even get it to do that in the laboratory.

Mentor
Klimatos, I understand that cold air flows down a slope. What I object to is your explanation of why. You give an incorrect impression that it flows against a pressure gradient, when in fact the flow is with and caused by a pressure gradient.

Gold Member
1)
2) I am more interested in explaining local weather phenomena than in global averages.

Addendum) I suspect that I am dealing with laboratory scientists here. The free atmosphere does not always behave as do samples of air in the laboratory. To me, atmospheric physics is primarily an observational science like astronomy, rather than a laboratory science like classical physics.
.

I realized that, which is why I have been stressing that it's the MEAN that counts (over the whole of the Earth). Unless you can come up with an explanation that introduces some form of artificial gravity, you can't suddenly have an area of higher pressure without, somewhere, having other area(s) with a lower pressure.
Your points about a vertical pressure gradient are not relevant either - there is always a vertical variation in pressure where there is gravity.

Classical Physics does pretty well in describing a whole lot of phenomena that are far too big to get in a Laboratory. Newton's (really ancient) laws give us a very good model for the Solar System, for example.

What is the location to do with the validity of Physics? Lab experiments are less complex and give you a chance of understanding, later, what goes on in complex situations. You have clearly been bamboozled by the sheer complexity of the weather and by a number of reddish herrings. They have, somehow convinced you that the basics of Physics don't apply. If the sum of all forces on the surface of the Earth doesn't equal the total weight of the material above it then there are either some 'sky hooks' or someone is squashing the atmosphere in a massive balloon.

I could, perhaps, agree that the ejecta from volcanic action could be steadily increasing the pressure due to the exit velocity. Alternatively, a slow stripping of the atmosphere by external effects could be removing atmosphere and causing a marginal reduction in overall pressure (conservation of momentum in both cases). But neither of those effects appear in your alternative model so we can ignore them for this purpose.

Mentor
Let us be careful here to talk about actual pressures (i. e., those measured by a manometer), not the fictional pressures that result when actual pressures are "reduced to sea level".
Ok then: consider a u-tube manometer with one tube open at the bottom of the Empire State Building and the other open at the top, on a calm day. What pressure does it read? Then take the same manometer to your glacier. Does the reading change?

Gold Member
Ok then: consider a u-tube manometer with one tube open at the bottom of the Empire State Building and the other open at the top, on a calm day. What pressure does it read? Then take the same manometer to your glacier. Does the reading change?

I think you could be wasting your time. klimatos is wedded to the idea that there is some magic at work due to the sheer complexity of the system.

klimatos
You are allowing the fact that a vertical pressure gradient exists in the atmosphere to confuse the issue.

The vertical pressure gradient IS the issue. The wind is blowing from the source area at a lower ambient pressure to the destination area at a higher ambient pressure. Ergo, it is blowing from a point of lower pressure to a point of higher pressure. Because there is a vertical pressure gradient, these winds are called katabatic winds; i. e. downslope winds. That vertical gradient is the essence of the issue.

A fluid can flow against the pressure force gradient whenever it is impelled to do so by a stronger force. In this case the force is that of gravity acting on a denser portion of the fluid. That is why these winds are often called "gravity winds". It is this gravity force that compels these winds to blow the way they do, not any pressure force. Quite the contrary, they are blowing against the pressure gradient.

In mountain areas, it is very common for such downslope flows to be isobaric. That is, there is no measurable pressure difference between the source area and the same elevation above the destination area. The same thing is very common with the ordinary alternation of land and sea breezes. The winds blow, impelled by density differences, despite the lack of any measurable pressure differences. Where then, does your "pressure gradient" come from?

Pressure differences are certainly the most important genesis of winds. But they are not the only such genesis. Density differences also play a role.

Gold Member
Yet again you are complicating the issue for no useful purpose. Take a body of air. It can be doing whatever you like within itself. If it is in equilibrium with its surroundings (not rising, falling or moving) then it is no more than a hot air balloon with no envelope. If you tell me that it is not in equilibrium, then choose a bigger body, which is. If you say that body has to be Earth sized, then so be it. The fact that it may be wrapped around a spheroid is not relevant, either. In the end, the total forces on it must be zero. All of the downwards (weight) forces must balance all of the upwards forces (caused by pressure on the ground. If this is not true then it must be accelerating upwards or downwards (outwards or inwards). If you want to introduce mountains, winds and anything else, you are just like the man in the van loaded with budgies. He can make no difference to the mean weight on his suspension.

Does Newton's First Law of motion mean nothing to you? Why should it not apply here?

No. Air cannot flow through an increasing pressure gradient.

There's a whole term for pressure gradients which are positive in the same direction as a flow: they're called "adverse" pressure gradients.

Klimatos, I understand that cold air flows down a slope. What I object to is your explanation of why. You give an incorrect impression that it flows against a pressure gradient, when in fact the flow is with and caused by a pressure gradient.

Is your impression that other common gravity currents (e.g., water flowing downhill) are caused by "favorable" pressure gradients?

Last edited:
Take a body of air. It can be doing whatever you like within itself. If it is in equilibrium with its surroundings (not rising, falling or moving) then it is no more than a hot air balloon with no envelope... In the end, the total forces on it must be zero. All of the downwards (weight) forces must balance all of the upwards forces (caused by pressure on the ground. If this is not true then it must be accelerating upwards or downwards (outwards or inwards).

That's a big if, around which the entire disagreement pretty much revolves.

The cold air flowing downslope in a katabatic wind isn't in equilibrium with its surroundings. That's why the katabatic wind happens.

Gold Member
That's a big if, around which the entire disagreement pretty much revolves.

The cold air flowing downslope in a katabatic wind isn't in equilibrium with its surroundings. That's why the katabatic wind happens.

That's why i said that you must take a bigger body - even the whole of the Earth's atmosphere, which is not moving, rising or falling. It's just easier to consider an ideal - bite sized bit - first. How about a nice stable cumulus cloud with the outside layers pretty well stationary and loads of turmoil inside it, sitting on a cushion of cool air?
However big you choose, I think you'd have to agree that Newton 1 must apply.

However big you choose, I think you'd have to agree that Newton 1 must apply.

But what does this have to do with the disagreement in the thread?

Gold Member
It resolves the disagreement about whether atmospheric pressure is due to the weight of the atmosphere. I must admit that I don't always read every post either. Haha

Mentor
Oy, this keeps getting worse. I need to go back and correct some errors right from the start, but first I am going to give another thought experiment that clearly illustrates the error (not that you responded to the last two...):

Imagine you are holding a cup of cold air and there is no wind to disturb it. The bottom of the cup is negligibly thin, but flat. There is pressure pushing up on the bottom of the cup due to the air around it and there is pressure pushing down on it due to the air above it. Are the pressures equal and if not, which is greater? Hint: punch a hole in the bottom of the cup and air will pour out as if it were water.

Mentor
Ok, first some definitions:

There are a few ways to measure pressure, but ultimately all pressure is measured as a differential, using a gauge. So differential pressure and gauge pressure measure the pressure difference between two different points. For absolute pressure, one of those points (the reference) is a vacuum.

In terms of the way pressure acts on things (a surface, other air), there are three possibilities: Static pressure is what you have when there is no movement. Dynamic (velocity) pressure is what you get due to movement, perpendicular to the movement of the air. Total pressure is the sum of these two.

So:
Atmospheric pressure is the pressure exerted on any exposed surface by the impact of air molecules upon that surface. It is the simple product of the number of impacts per square meter per second and the mean impulse per impact.
No. That description includes velocity pressure. Atmospheric pressure is the static pressure of the atmosphere. Mistakenly capturing velocity pressure in the measurement would crash airplanes.
When the wind is blowing, that approximation becomes increasingly less accurate.
No. The wind does not affect the approximation, just the measurement accuracy of certain instruments.
If the wind is blowing then it must be from a higher pressure region to a lower pressure region.
Important point of clarification. We vs klimatos are talking past each other here a little bit, but only because the typical description above is usually sufficient. But when those "regions" become separated vertically, it isn't quite right anymore. That's not your fault, as I know that's what you meant: klimatos threw a non sequitur into the mix that confuses the issue.
There are lots of winds that do not blow from areas of high pressure to areas of low pressure. Virtually all gravity winds blow from areas of low pressure to areas of high pressure. If you think of winds as three-dimensional phenomena (the only rational perspective, in my opinion) then all areas when the air is sinking represent movement from low pressure areas to high pressure areas.
So these "vertical winds" break the simplification. But:
No. Air cannot flow through an increasing pressure gradient. Gravity causes the pressure gradient that moves the air in the case of gravity winds.
Pressure gradient. I've clarified the situation to be more precise. You then later misused the term, so I was right that you don't realize what it means and why the distinction is important. A gradient is a vector field showing the variations in a scalar field. In other words, the difference between two widely spaced areas is not a gradient: a gradient only exists at a point. If the pressure variation between two widely spaced areas were continuous, then you could average it and get a gradient at every point, but that isn't the case here and that's where your error lies. As my cup example demonstrates, there is a discontinuity in the pressures. The pressure just above the bottom of the cup is higher than the pressure just below the bottom of the cup, so when you punch a hole in the cup, the gradient at the point between them is high->low is above->below. So air spills out of the cup due from an "area" of higher pressure to an "area" of lower pressure despite the fact that two widely separated parcels of air will appear to show an opposite pressure gradient.

So how do we explain this or expand it more generally. Well consider that the air spills into another cup below. As it spills into the cup below, the higher ambient pressure compresses it, making it more dense and still a higher pressure than the air around it. So essentially, the falling parcel of cold air carries with it, follows or creates its own pressure gradient as it falls.
Under isobaric conditions, cold air will flow towards warm air. At 1000 hPa, the flux rate for air at -25°C is 3.11 x 10^27 molecules per square meter per second. At +25°C, the flux is 2.84 x 10^27. This flux differential will manifest itself in a flow of air from the colder to the warmer. This is the genesis of the very common daily alternations of "land breezes" and "sea breezes".
No. There can't not be a pressure difference between warm and cold parcels of air next to each other: the cold parcel is heavier, so its pressure is greater. Land and sea breezes are and example of this, not a counterexample.
Finally, there will be a net flow of air away from evaporating surfaces and toward condensing surfaces.

These are all common illustrations of isobaric winds.
Evaporation causes local pressure differences.
Every time there is subsidence in the atmosphere (the downward leg of a Hadley Cell, for instance) you have air moving from an area of lower real pressures (not pressure reduced to sea level) to an area of higher real pressures.
It's not pressure reduced to sea level that matters here, it is absolute pressure or pressure differential: pressure with the normal atmospheric pressure gradients removed.
The vertical pressure gradient IS the issue. The wind is blowing from the source area at a lower ambient pressure to the destination area at a higher ambient pressure. Ergo, it is blowing from a point of lower pressure to a point of higher pressure. Because there is a vertical pressure gradient, these winds are called katabatic winds; i. e. downslope winds. That vertical gradient is the essence of the issue.
No. It's a non-sequitur that you've added, which is confusing you and the issue.

Pressure gradient... A gradient is a vector field showing the variations in a scalar field. In other words, the difference between two widely spaced areas is not a gradient: a gradient only exists at a point. If the pressure variation between two widely spaced areas were continuous, then you could average it and get a gradient at every point, but that isn't the case here and that's where your error lies.

An exceedingly small time after you opened the hole in your coffee cup, there would no longer be a pressure discontinuity but only a very steep gradient, and yet the flow would continue for a while longer. Hence the discontinuity is not really the sticking point.

The pressure just above the bottom of the cup is higher than the pressure just below the bottom of the cup, so when you punch a hole in the cup, the gradient at the point between them is high->low is above->below. So air spills out of the cup due from an "area" of higher pressure to an "area" of lower pressure despite the fact that two widely separated parcels of air will appear to show an opposite pressure gradient.

This thought experiment seems a bit misleading as described.

You could very well imagine the same experiment, but with water in the cup. Most observers would agree that the water at the bottom of the cup is at higher pressure than the air at the same height outside the cup, and thus the water would be pushed out under some pressure. However, the "last" parcel of water would still fall out of the hole (let's assume a large hole here), even with no additional weight of water above it. This would happen solely due to the weight of the parcel itself, which is heavier than air.

klimatos
This thread has gone badly off post. The OP asked the question of why atmospheric pressures were omni-directional when the weight-force hypothesis used to explain such pressures was uni-directional. That question has been answered; and the discussion has since fragmented.

I intend to abandon this thread and initiate a new thread on the most interesting of these fragments: the question of whether fluids can naturally flow from areas of lower pressure to areas of higher pressure.

Please join me there if the subject interests you.