Skip to comments.The Perspective of a Lifetime on Atmospheric Modeling
Posted on 08/04/2012 11:57:56 PM PDT by neverdem
The president vowed to make climate change a top priority in his second term, suggesting that a major assault on industry is coming if he is re-elected. So before the potential onslaught, some real-world perspective on climate change is essential.
First, note that the tool used to both develop future global climate scenarios and to panic the public on meteorological mayhem is atmospheric modeling.
Most of my nearly 35 years of professional life has been involved with atmospheric modeling in one...
What I and so many other air modelers have discovered is that, as impressive as modeling has become, model results beyond the immediate downwind distance of the pollution source and within a relatively brief amount of time, are not very reliable, despite the awesome computing power available today. We know that dependence on their output is quite limited and to extrapolate too far beyond the bounds of the model assumptions is foolhardy.
Compare the experience of thousands of non-academic air modelers with the largely academic and government climate modelers. Their combined efforts have produced impressive results in scope and scale, yet, like air pollution modeling, their model outputs for long-term global climate conditions still boil down to limited guesses.
Regardless, a bit of understanding about the global atmosphere has been spun into a trillion-dollar bonanza by a colaition of supporters. These cheerleaders take the form of career politicians, bureaucrats, environmental and social activists, academics and educators, technologists and consultants, journalists, bloggers, and groupies of all stripes.
But realism and humility about the limitations of climate modeling must set in soon with enough scientists and those of the general public who care enough to pay attention. Let's face it: if Mr. Obama gets his way in November, then more than our supposed climate future with be in dire straits.
(Excerpt) Read more at americanthinker.com ...
The Navier Stokes equations are differential equations that describe the flow of fluids with changes in density and temperature. They are used to describe weather or climate. They are non-linear, chaotic, with sensitive dependence on initial conditions.
That means, no finite description of a starting state, or any finite description of a number of starting states are sufficient to enable prediction of a long term future state.
This has been known since the 1963 paper by Edward Lorenz “Deterministic Aperiodic Flow”. It is a founding work in the study of chaos.
It took about 24 hours of computation time on a very serious supercomputer.
The world is a very big place, the vary few climate interaction sub models that have been characterized are also very non liner.
It's really tough to say about the interactions between the multitude of the climate interactions mechanisms that we don't yet realize exist along with the other mechanisms that we have a hazy realization of but have no realistic characterization or model pretty much makes any realistic analysis intractable.
I am not quite as arrogant as the Global Warming gurus so I have to admit I have absolutely no idea what realistic boundary conditions on a world wide scale might be or how to properly apply them. Given the complexity and scope of the problem and enormous size of the world, a slightly more humble perspective might be in order for some of the more vocal researchers.
The biggest factors are the amount of solar radiation energy that hits the earth and how much of that energy is either radiated or reflected back into space.
The amount of solar energy from the sun hitting the earth is constantly changing, we have no control over the sun, and very little ability to accurately predict what the sun's output will be over any extended length of time.
The amount of the sun's energy the Earth radiates and reflects back into space are not constants and the parameters that govern the reflection and radiation have not been sufficiently characterized to provide the input for an accurate, and possibly not even realistic basis for climate analysis.
Having tried my hand at Navier Stokes modeling in the past, let me add to what you said. As bad as Navier Stokes is to model, they are just a subset of the equations needed to do climate modeling. There are also chemical, phase shift, biological, etc. factors not even considered.
In other words, the only way for these computer models to be accurate to the degree these “environmentalists” claim, is for them to start by knowing the exact energy vector of every molecule of everything everywhere.
Translated into English, for those who don't speak PDE, these equations cannot predict how large a temperature change will result from very small changes in CO2, or even the direction of change. In chaos theory, the butterfly effect refers to a hurricane's formation depending on how many times a distant butterfly had flapped its wings several weeks before. In the correct application of math to global warming theory, the nature of the equations means that the extra CO2 emitted by a butterfly can change whether the earth warms or cools and can change that outcome in either direction.
There are also chemical, phase shift, biological, etc. factors not even considered or even known.
Having experience with clinical studies and controlled laboratory experimentation, I’ve always wondered how any advocate of anthropogenic global warming can even begin to make the claim that a sufficient number of climate data points have been collected with which to predict future trends. Apparently, plugging a few measurements into a supercomputer and running them through differential analysis is supposed to make up for the fact that there simply are not sufficient data points collected over a long enough period of time to even begin to derive appropriate equations for modeling long-term climate changes.
Models are dress-rehearsals at best. Any similarity to reality is either coincidental, temporary, or both.
With media and politicians seeing your studies as a tool to increase their power and influence, they will assist you in flat out making _hit up. Who needs data points?
You just tell the sheeple how many millions of data points billions of calculations and how difficult it is to grasp and understand the science and 90% of the people will believe.
The problem for the Global Warming gurus is that humble people do not get grants.
People that admit that they are just whistling in the dark dont get grants. Unless you can tell some bureaucrat that you can produce tangible and useful results you dont get a grant.
Given the over production of people with degrees in the last few decades the competition for a paying job in science is getting fierce. These people want money and the big money is in scaring people. For climatologist the easy money is in scaring people with Global Warming.
This is correct. The other problem is entering these vectors into the absolutely correct computer model (which does not exist) to infinite precision (which does not exist).
This pretty much sums up the world-is-ending, humans-are-evil, surrender-to-the-world-bureaucracy, Chicken-Little Climate Crisis.
All models are inaccurate, though occasionally, some models are useful. In general, though, the more complete a model is, the less useful it becomes. This is due to the difficulty of getting accurate input data for the model, and computing capacity to run it.
Some noteworthy articles about politics, foreign or military affairs, IMHO, FReepmail me if you want on or off my list.
I wish the same level of effort went into creating supercomputer economic models, the results of which would do more good for mankind. The models could be turned into a competitive game for mass consumption to teach the damaging effects of carbon taxes, regulation, and welfare spending. A problem though is the left would use these same models to study how most efficiently to destroy the economy.
Not good enough. Remember at that point you have to calculate state-to-state energy transfer and reaction cross sections (which are inherently probabilistic), and then worry about chaotic effects...
In other words, fuggetaboutit.