The American Meteorological Society is soliciting input on their draft statement (h/t to Joe Daleo)
They have sent out the following
The following draft statement is currently under review by the AMS Council: “Communicating Science”
If you have comment on this draft AMS Statement currently under consideration, you may transmit those comments to the AMS Council by sending a message to the following e-mail address:
email@example.com_ (mailto:firstname.lastname@example.org) before 2 February 2011.
Melissa S. Weston, Executive Officer
American Meteorological Society
45 Beacon Street
Boston, Massachusetts 02108-3693 USA
Web: _http://www.ametsoc.org_ (http://www.ametsoc.org/)
I urge readers of my weblog to send in comments. I have excepted a few statements from the draft text for my comments.
The first excerpt that I am commenting on is
“What Is Science?
Science is an enterprise that systematically acquires and organizes knowledge in the form of testable explanations and verifiable predictions about the natural world.”
This is an excellent succinct summary statement of what science is. This is clearly a requirement that multi-decadal climate predictions for the remainder of the 21st century cannot satisfy.
The second excerpt is
“A model is a physical, mathematical, or conceptual framework for describing reality. Most weather and climate models, for example, are sets of equations that represent the principles and conditions that govern the behavior of energy, mass, momentum, and moisture in the atmosphere.”
The statement needs to have added that only the dynamic core of these models (i.e. the pressure gradient force, advection, and gravity) represent basic physics. All other components of these models are parameterizations , which means they are not fundamental physics but engineering code with tunable parameters.
“Modeling will continue to be an essential tool used by scientists for investigation and prediction. Because models are solidly grounded in mathematics and physical observations, and represent a logical description of the system that they are designed to predict, they offer the best and most effective means for testing the consequence of a full range of inputs to a system, from subtle changes to shocks.”
This statement has an important error. Modeling DO NOT “offer the best and most effective means for testing the consequence of a full range of inputs to a system, from subtle changes to shocks”. Models are themselves hypotheses. They cannot be used as a “test” without comparing their predictions with actual observed data.
Real world observations, not models, therefore, are the appropriate means for testing the consequence of a full range of inputs to a system, from subtle changes to shocks! Models can assist in the interpretation of the behavior of the studied system, as well as to make forecasts ONCE the accuracy of their predictions are verified against observed data.
This is clearly a requirement that multi-decadal climate predictions decades from now cannot pass until those decades have occurred.
I discuss this issue in a number of my posts; e.g.
The bottom line message is the weather and climate models are not basic physics, but are engineering codes with a core of fundamental physics but with much of the atmospheric, ocean, cryosphere and land represented by tunable engineering code (i.e. their parameterizations).
I urge the AMS statement be modified to correct this misunderstanding concerning models.