Environment = Sum(Local) + Sum(Global)

[petrushkagoogol]

I prefer to divide natural phenomena into 2 categories : local and global

 

For instance, an ocean is global whereas a desert is local.

 

A cyclone is global whereas a hailstorm is local.

 

A mountain is local and so is a lake. and so on.

 

At any point in time the sum total of local and global factors defines our immediate weather condition.

 

Both are contributors to the micro environment and coexist.

 

Better understanding of both are required to give an accurate forecast of environmental hazards.

 

We should change our approach to weather forecasting to be more holistic as a result.

 

[CraigD]

We should change our approach to weather forecasting to be more holistic as a result.

What (links and references, please!) makes you think weather modelers have not been trying to do this for many decades? What, specifically, do you believe needs to be changed in existing climate and weather modeling techniques to improve them?

 

I’m not a climate or weather modeler, nor even a follower of their literature, so can’t claim with any confidence that they should do or not do anything differently.

 

However, the suggestion that the division between small-scale, local models, and large-scale, global one, is useful, doesn’t seem correct to me, because of ideas I was introduced to in science historian James Gleick’s 1987 book Chaos: Making a New Science.

 

In it, Gleick described meteorologist Edward Lorenz as “the first true experimenter in chaos.” In 1960, Lorenz was one of the few meteorologists working on computer programs to predict the weather. This was not because meteorologists weren’t interested in the subject, but because computers and skill in using them were rare then.

 

Lorenz had high and justified expectations that, as the cost and performance of computers increased, he and others would be successful in using computers to predict weather years in the future with tremendous accuracy. Building on work by Lewis Fry Richardson, such computer luminaries as John Von Neumann had from the 1920s done work and estimates suggesting that all that was needed for computer programs like this to succeed were faster, larger memory computers. Lorenz was part of an optimistic continuation of that work.

 

What Lorenz discovered was that a key assumption underlying this expectation was wrong. This assumption is that small differences in the initial data in such models result in small differences in the data later in the simulation. Instead, he found that small differences led to very large differences. This became know as the butterfly effect (Lorenz first called it “the sea gull effect”, but “butterfly” was judged more poetic and just as accurate), which mathematician and science writer Ian Steward described in his 1989 Does God Play Dice: The New Mathematics of Chaos:

The flapping of a single butterfly's wing today produces a tiny change in the state of the atmosphere. Over a period of time, what the atmosphere actually does diverges from what it would have done. So, in a month's time, a tornado that would have devastated the Indonesian coast doesn't happen. Or maybe one that wasn't going to happen, does.

It’s critical to understand that this is not a metaphor, or hyperbole. It’s a physically real prediction, supported by computer simulations shown to be sufficiently accurate to predict tornados or the lack of them a month in the future. A difference in the Earth’s atmosphere that could be caused by a butterfly, anywhere on Earth, flapping or not flapping its wings at a specific instant results in a future where a tornado does or doesn’t occur.

 

It’s also critical to understand that the butterfly effect doesn’t imply that a the wing flapping of a butterfly, sea gull, or the detonation of an arsenal of thermonuclear bombs can cause the oceans to evaporate or freeze over at some distant point in the future. Chaos doesn’t make it impossible to predict the range of possible future weathers, just to predict them precisely.

 

It doesn’t matter what concepts you use – local, global, or other – to build your weather modeling program. Chaos assures that you’ll never be able to measure initial conditions accurately enough to predict a specific tornado months or years in advance.

 

A good history of computer weather prediction can be read in P. Lynch, The origins of computer weather prediction and climate modeling, J. Comput. Phys. (2007), doi:10.1016/j.jcp.2007.02.034.

[petrushkagoogol]

What (links and references, please!) makes you think weather modelers have not been trying to do this for many decades? What, specifically, do you believe needs to be changed in existing climate and weather modeling techniques to improve them?

 

I’m not a climate or weather modeler, nor even a follower of their literature, so can’t claim with any confidence that they should do or not do anything differently.

 

However, the suggestion that the division between small-scale, local models, and large-scale, global one, is useful, doesn’t seem correct to me, because of ideas I was introduced to in science historian James Gleick’s 1987 book Chaos: Making a New Science.

 

In it, Gleick described meteorologist Edward Lorenz as “the first true experimenter in chaos.” In 1960, Lorenz was one of the few meteorologists working on computer programs to predict the weather. This was not because meteorologists weren’t interested in the subject, but because computers and skill in using them were rare then.

 

Lorenz had high and justified expectations that, as the cost and performance of computers increased, he and others would be successful in using computers to predict weather years in the future with tremendous accuracy. Building on work by Lewis Fry Richardson, such computer luminaries as John Von Neumann had from the 1920s done work and estimates suggesting that all that was needed for computer programs like this to succeed were faster, larger memory computers. Lorenz was part of an optimistic continuation of that work.

 

What Lorenz discovered was that a key assumption underlying this expectation was wrong. This assumption is that small differences in the initial data in such models result in small differences in the data later in the simulation. Instead, he found that small differences led to very large differences. This became know as the butterfly effect (Lorenz first called it “the sea gull effect”, but “butterfly” was judged more poetic and just as accurate), which mathematician and science writer Ian Steward described in his 1989 Does God Play Dice: The New Mathematics of Chaos:

The flapping of a single butterfly's wing today produces a tiny change in the state of the atmosphere. Over a period of time, what the atmosphere actually does diverges from what it would have done. So, in a month's time, a tornado that would have devastated the Indonesian coast doesn't happen. Or maybe one that wasn't going to happen, does.

It’s critical to understand that this is not a metaphor, or hyperbole. It’s a physically real prediction, supported by computer simulations shown to be sufficiently accurate to predict tornados or the lack of them a month in the future. A difference in the Earth’s atmosphere that could be caused by a butterfly, anywhere on Earth, flapping or not flapping its wings at a specific instant results in a future where a tornado does or doesn’t occur.

 

It’s also critical to understand that the butterfly effect doesn’t imply that a the wing flapping of a butterfly, sea gull, or the detonation of an arsenal of thermonuclear bombs can cause the oceans to evaporate or freeze over at some distant point in the future. Chaos doesn’t make it impossible to predict the range of possible future weathers, just to predict them precisely.

 

It doesn’t matter what concepts you use – local, global, or other – to build your weather modeling program. Chaos assures that you’ll never be able to measure initial conditions accurately enough to predict a specific tornado months or years in advance.

 

A good history of computer weather prediction can be read in P. Lynch, The origins of computer weather prediction and climate modeling, J. Comput. Phys. (2007), doi:10.1016/j.jcp.2007.02.034.

 

 

I am not fully acquainted with chaos theory but the inverse appears realistic - big affecting small seems more realistic than a butterfly causing a typhoon... :-)