The weather.

[peter]

Not a climate change thread - simply a weather thread.

My wife and I were out walking yesterday and the wind was all over the place. Gusting here and then there, stopping briefly and then catching your hat again just as you relaxed your grip, hitting you from one side - then the other. I said to my wife "Why does it do that". "What?" she replied. "Jump about all over the place," I said. "I mean, I get it if it wants to go from A to B, like hot air over the land rising and the cool air from the sea rushing in to fill the gap [even out the pressure if you like] - but whats with all this 'all over the place' stuff'".

She didn't know the answer and I guessed that it was something to do with micro variations in the various heat retaining capacities of different bits of land/sea surface creating a chaotic system, resulting in a chaotic wind. If I was a scientist I guess I'd devise an experiment and go out and test this theory.

two things arise from all of this; i) whether my hypothesis was correct or bullshit and ii) whether this is the way science is done - and if so, what have I just described, the inductive or the empirical approach?

[Hashi Lebwohl]

Turbulence. The atmosphere is a continuous body of air but there are pockets inside it and these pockets are essentially independent of one another; the places where these pockets interact are what we call fronts and they give rise to most active weather conditions. Pockets of air interacting create turbulence and that turbulence manifests on the ground as what we feel as gusty wind blowing around. In more intense cases that turbulence coalesces into tornadoes.

In urban areas we feel gusty wind coming at us from different angles because of buildings. Wind current higher up off the ground hit the buildings, get directed downwards, and blast us on the street. If one current hits two or three buildings we will get downdrafts from multiple directions.

Many people underestimate the power of wind. I saw a documentary clip a couple of years ago where one researcher thinks that some ancient megalithic structures might have been raised off the ground by attaching large kites to ropes wrapped around the object needing to be lifted. A sustained 50 kph wind is sufficient to help lift stone objects that twenty people would have difficulty lifting via an a-frame block-and-tackle pulley system. Fascinating stuff. The only problem with that hypothesis is that we haven't seen any hieroglyphs depicting it, as far as I know, and Ancient Egyptians depicted many mundane, everyday activities in that manner.

[peter]

But what actually causes these independant pockets Hashi - given the short duration of the 'blasts' as they were striking us they must have been relatively small in size. We were in an area with trees - but in no sense thickly wooded, and while there was certainly a prevailing ditection to the wind it was tubulent to a degree that I find quite difficult to understand. I guess you only have to look at water tumbling over a waterfall to see how chaotic these systems are - but in the waterfall we have the stones and rocks of the river-bed to explain the resultant malestrom; what are there equivalent in the weather systems such as the one I walked in yesterday?

[Hashi Lebwohl]

Pockets develop in the atmosphere as a side-effect of being such a large body of relatively compressed gas in a non-thermodynamically-stable system. If you fill a large jar full of air and put it on a shelf you will never notice any circulation in that system--the gases remain static, form a macro point of view. Put a small source of heat on one side, though, and the hot air rising begins to set up an internal system of circulation even if you can't see it.
Air sitting over a body of water on a hot day becomes denser as water vapor evaporates into it. This warm, dense air might be sitting next to its neighbor which is cooler and/or less dense so it will start to "leak" under that neighbor. Where these two pockets interact will give rise to winds and possibly rain or, in more extreme cases, thunderstorms. In short, the pockets arise from the conditions in their locales.

Some weather systems also include mesocyclones, areas of rotating winds that may form tornadoes but often do not. These mesocyclones also create downdrafts, bursts of winds that rush out of the cloud layer, slam into the ground, bounce off, and become straight-line winds along the ground.

Turbulence is its own subcatagory of fluid dynamics, a difficult science to study that is very math-intensive. Even most university-level science majors never study it, unless they are majoring in physics with a focus in dynamics.

[Vraith]

But what actually causes these independant pockets

Many things---land contours and differential heat are the big ones.
But it also relates to velocity...high enough speed will make it tend towards laminar.
It's kinda thousands of small-scale butterfly effects.

IIRC, even across perfectly flat surface, and consistent temperature, there would be turbulent and laminar flows---depending on velocity and viscosity and density.
And---IIRC---the turbulence has an ideal range. For a given density and viscosity, the flow will be smooth below a specific velocity, and above a specific velocity and turbulent between those two velocities.
[[certain soaring birds---almost positive the albtross, maybe others---actually USE the turbulence for extremely energy-efficient, very long-range gliding...laminar flow and ordinary lift just won't do the job]]

[peter]

Both of the above replies seem to imply that the prospects for long-range weather forcasting are limited in the extreme. This system of butterfly effect within butterfly effect within butterfly effect would I imagine, while in theory be possible to model, in practice be as close to impossible as you can get. I am told by farmer aquaintances I know, that weather forcasting is pretty accurate these days up to a week or even 10 days in advance [which seems a mighty hike from the two or three days accuracy of my youth] - but it seems unlikely, given the inherently chaotic mechanism[s] you have outlined above, that the process will be able to be extrapolated much further [since the effect of minor innacuracies would be exponentially magnified with each sucessive days prediction would they not?].

[Hashi Lebwohl]

General circumstances can be reasonably accurate up to a week out but actual weather events are essentially impossible to predict or model more than 3 days out. Meteorologists might know that conditions are favorable for thunderstorms in the Metroplex today--which they are--but even the best models cannot tell exactly when or where an individual storm will develop and they cannot tell how strong the storm may be. There is still a lot of uncertainty and guesswork in meteorology...except after the fact when they are studying the data from a weather event.

[Vraith]

Both of the above replies seem to imply that the prospects for long-range weather forcasting are limited in the extreme.

Yes, pretty limited. Depending on the scale and level of detail.
There is uncertainty, as Hashi pointed out, that will never vanish---it will always be nothing but probabilities...though we can refine them, as we gather constantly more massive quantities of data, better models, and faster computers.
But that's how many things work.
As it stands now, the CDC or a major university or Health Insurance number-cruncher could tell you, with nearly 100% accuracy/near zero margin of error, how many people in the U.S. and/or EU will die of a heart attack tomorrow.
But there is an extremely low probability that anyone can identify any particular person who will die tomorrow.
Last winter, people bitched and whined about preparing for a monster East Coast storm that never happened...but they're wrong/silly, and don't realize how lucky they were. The forecast was "wrong" by much less than 1%...which was just enough that the tuna and killer whales got the bad weather instead of the people.

[Hashi Lebwohl]

As it stands now, the CDC or a major university or Health Insurance number-cruncher could tell you, with nearly 100% accuracy/near zero margin of error, how many people in the U.S. and/or EU will die of a heart attack tomorrow.
But there is an extremely low probability that anyone can identify any particular person who will die tomorrow.

That reminds me of an old joke.

Q: What is the difference between an actuary and a Mafia actuary?

A: An actuary can tell you how many people are going to die this year. A Mafia actuary can name them.

[peter]

Any observations on how many days in advance the current forecasts can go with say, 80% accuracy?

[Hashi Lebwohl]

I don't have hard figures with which to verify this but I usually presume about three. Anything farther into the future than that and you may safely ignore it as mere guesswork.

[Vraith]

I don't have hard figures with which to verify this but I usually presume about three. Anything farther into the future than that and you may safely ignore it as mere guesswork.

I can't find it now, but sometime in the last few months ran across a "most unpredictable weather" thing. [The most unpredictable U.S. city was in Iowa or Utah or somesuch].

You're part of the world, IIRC, is a mix...Houston/coast is fairly predictable, Dallas area and all up into Oklahoma and points north is a bit toward the unpredictable side [because of tornado season, mostly---skews the rest of the year].

But places like Phoenix, and the lower third of California...you could plan your wedding today for a date in 2020, and be pretty confidant of the weather. [the entire West Coast was pretty predictable, but that southern section practically clockwork.

[Hashi Lebwohl]

Dallas area and all up into Oklahoma and points north is a bit toward the unpredictable side

I live this reality.

[peter]

Where I live [South-West England] we joke that you never get more than three days of weather the same - but it's actually pretty true. Our entire weather forecast can be summed up in one word; Changeable!

[Vraith]

Where I live [South-West England] we joke that you never get more than three days of weather the same - but it's actually pretty true. Our entire weather forecast can be summed up in one word; Changeable!

But---is it only changeable, or is it also unpredictable? South-West...wouldn't that be pretty close to the sea [not that any of England is really far from it]?
Coastlines [at least for the U.S. analysis, which I'm assuming applies generally] can be pretty changeable---but are also [generally] very predictable. [It may be a rain of frogs in the A.M. and beach day after noon---but you can see it coming].

[michaelm]

Where I live [South-West England] we joke that you never get more than three days of weather the same - but it's actually pretty true. Our entire weather forecast can be summed up in one word; Changeable!

When I lived in New Hampshire it was similar - you just didn't know what you were going to get next. It could go from sunny to rainy to windy to snowy over consecutive days, and you could get a few warm days in winter or is could snow in May.

Of all the places I've spent time in I think the New England area has the most changeable weather.

[peter]

Vraith has a fair point there; changeable it certainly is, but also fairly predictable. The forecasts certainly seem good for a couple of days [which is about as far ahead as I ever think ahead anyways ].

Is New England named for it's similarity to England in terms of appearance alone or in general weather conditions as well {I assume the second has significant influence on the first in any case} or just coz lots of English settlers made their homes their. Pictures of the famed New England 'Fall' look absolutely stunning. I'd love to see that one day.

[Vraith]

Is New England named for it's similarity to England in terms of appearance alone or in general weather conditions as well

I think it's mostly because of the settlers, but that's just an assumption. I bet WF knows, I think he's a life-long New Englander, whereas I only spent a few years there...you should look at a map at a scale with town names, compare U.K./England to New England, especially coastal. Some of it is practically mirror image.

There is some weather commonality, again especially coastal. Cuz near the sea, and the Gulf Stream and such.
But also some real differences because of the general massive flow in New England that crosses a continent/land [part of it semi-polar] and moves offshore, whereas England it crosses an ocean and moves on-shore.