Philosophical reflections on climate model projections
Posted on August 12, 2012
by Judith Curry
Should probabilistic qualities be assigned to climate model projections?
Are the approaches used by the IPCC for assessing climate model projection quality – confidence building, subjective Bayesian, and likelihood – appropriate for climate models?
What are some other approaches that could be used?
“The basic theory of a climate model can be formulated using equations for the time derivatives of the model’s state vector variables, xi, i = 1, …, n, as is schematically represented by … Eqt. (1) . In Eqt. (1), t denotes time, the functions Gi represent external forcing factors and how these function together to change the state vector quantities, and the Fi represent the many physical, chemical and biological factors in the climate system and how these function together to change the state vector quantities.”
Moderation note: This is a technical thread and comment will be moderated for relevance.
Reading the terms and claims in this thread beginning it confused me in my understanding of philosophy, techniques, and software modeling of thermal physical processes in geometric structures.
Physical processes are described in equations using well defined elements and their dimensions. Equations showing the respect to the science of philosophy in that it is recognized that nothing can be true and untrue at the same time.
In the case of thermal physics this means that the physical process is described as a heat current floating from warm to cold, and is generated by a heat source in Watt. At each location in the circuit the temperature can be calculated depending from the impedances and thermal resistors and the moving fluids in the geometry as function of time, if there are changing elements as well changing heat powers over time.
In the case of modeling the heat current from a source to the Earth and further into the cold space this means that the model must be constructed from the real elements and the real geometry.
It is an old good tradition in physics to accept values of elements which are not known accurate, but it is not allowed to bias a model with belief functions of a magician or an authority. It’s not science.
It may be a genius work to describe the thermal physics of global heat in one equation, but this is not necessary, because other mathematical models using iteration steps and a comparison of the global temperatures over the past with the calculated temperatures over the past from the model. In such a model the model parameters have to change as long as the global temperatures over the past fit with the used model parameter to a minimum.
It is clear that in such a model with a lot of temperature frequencies there must be defined appropriate oscillators, which build the known reconstructed temperature spectra from the literature.
This was also done to make precise predictions of the tides:
“The IUGG (international union for geodesy and geophysics) called an international working group in the year 1965 from mathematicians to assistance, who had come however after 10 years work to no solution contently placing. Their mathematical models provided for example for the North pacific ebbs-tide ahead although floods were observed and turned around. In the year 1972 U.S. of satellite and/or rocket designs required a forecast of the Tide height on 10 cm exactly. After 6 years modeling time appeared the North pacific in the spring 1978 then in tidal situation true to nature. For this model moon and sun became mathematical because of their elliptical and inclined orbits by a row of fictitious moons and suns replaced. For an accuracy of 10 cm to reach, they needed 6 moons and 5 suns with 4 halve a day’s, 4 complete days and three longer periods (14 days, month, and halve a year). After further 6 years the model was then extended of the North pacific.”
It is out of question that the oscillation frequencies of the global temperature spectra have physical causes. But as long as the source of these oscillators are unknown a model can be build which make use of it. Dynamic impedance in the heat current as an effect of known volcano events van be added to the model as a function of time. The first step can be reached to simulate the global temperature precise as reconstructed from the proxies.
There are powerful FEM software tools, which can solve thermal processes and fluid processes in multitasking. An example I can show is the temperature in Kelvin of 3D geometry in that a heat is loaded in a volume and the heat is streaming in a colder floating fluid to the right:
http://www.volker-doormann.org/images/heat_flow.jpg
I think it is not out of the question that in a second step the nature of the used and needed oscillators and its strengths to simulate the reconstructed global temperatures can be found.
In general I think it is a method of science to find relations of physical processes, based on geometries and energies, and it is not an appropriate method in science to start with a mathematical equation of vectors and ‘factors’ without any physical dimensions, because it has no basis in physics and not in the very real thermal nature of the Earth temperature spectra with frequencies of decades of kiloyears^-1 to month^-1.
V.