Why Climate Modeling Cannot Work - Page 2 - New England Patriots Forums

wistahpatsfan · 06-14-2011, 04:53 PM

Quote:

Originally Posted by patsfan13

Interesting observations wistah. The reason Brute computing power cannot solve the problem has to do with the underlying nature of the systems. The sensitivity to initial conditions and the propagation of small errors through the system make exact predictions impossible over a long time series as I read it.

That does not mean that we can't say for example that it will tend to be colder in Jan that in July. What we can't say is that on Next Jan 23 in Boston there will be 10" on snow the wind will avg 15mpg and the high temp will be 23F.

Think of a river, we can determine that the flow will go from a higher potential to a lower potential (ie water will flow downhill, but to try to predict the behavior of a group of mocules and the behavior of the eddies, and rapids is beyond our scope to pin down specifically.

Many natural systems are chaotic those are the kinds of problems science has a hard time pinning down.

If you area about 'attractors' in chaos theory, you will see a mechanism that shows up a lot in nature (and electronics).

I knowq a lot about watershed modeling. Very rarely is it necessary to include the characteristics of eddys and pebbles in potholes. As far as humans are concerned in many cases, what matters is knowing the big picture, and that is how high will the river be 10 miles downstream from a rain event and when it will happen. Chemical compositions and temperature are also important. Other factors are assessed and inserted into the model such as turbuduty, stream contours, gradient, and local surface conditions like bank vegetation, surface permeability, soil characteristics and bedrock countours. There are many dynamic and ever-evolving models that can be used for these assessments. Thing is, until about 30 years ago, watershed science and management was in its infancy as a discipline and models were crude to say the least. They aren't anymore. And while the small pictures are important to some on an event-to event basis, policy making is more dependent on on-going improvement of the assessment tools and should be made to the best of our ability until the science and modeling improves, at which point we can and should make adjustments.