A Complex Answer

This blog is meant to be an antidote to the RealClimate blog, but it seems rather churlish of me to attack that blog for the same reason as that given by George Mallory when asked 'Why climb Everest?'. He replied "Because it is there". So, when Spencer R. Weart guest posts "Simple Question, Simple Answer… Not" which is a thinnly veiled attack on we engineers (I have a BSc in Electrical Engineering), how can I refuse to respond :-)

Spencer claims that we engineers want a simple answer to global warming, but in truth what we want is an answer that works. What the climate modelers fail to see is that a computer model is not a valid proof that a theory is correct. For two millenia Aristotle's ideas were held sacrosanct until Galileo used experiments to show them wrong. It was this empirical breakthrough that led to the scientific revolution. Now Syukuro Manabe reputation in climate models is held in the same regard as was Aristole in natural science. When Christy and Spencer showed that his model is wrong, they like Galileo became scientific outcasts.

The problem with the Manabe-Wetherald system is not that it is too complex for engineers to follow, but that it is too simple to match the complex earth system. It is based on the idea that as the Earth's surface warms and more infra-red radiation is emitted, then the proportion of this radiation escaping to space will match the incoming solar radiation. Plausible, but wrong!

The outgoing radiation to space, in the greenhouse bands, is emitted by the atmosphere not the surface. This is obviously true for ozone, but is also true for carbon dioxide. The surface of the atmosphere that radiates to space is isolated from the surface by a thick layer of the atmosphere in local thermodynamic equilibrium. If you increase the CO2 concentration you do close the IR window but that effect is trivial, because the main CO2 band around 667 cm-1 is already saturated.

The balance between incoming solar radiation (ISR) and outgoing longwave radiation (OLR) which must exist at the top of the atmosphere only, is not maintained by variations of OLR, but by changes to the ISR. On Earth the ISR, which depends on the plantary albedo, is mainly controlled by water clouds. When the surface temperature rises more of it is vapourised and the cloud cover becomes thicker. On Venus the same effct happens, but there it is surphur that is vaporised to form the thick SO2 clouds of Venus. The high surface temperature of Venus can be explained by its need to reach the melting point of sulphur. The Martian temperature is limited by the formation of dust clouds.

On Earth (and Mars) there is an additional factor affecting albedo - surface ice. In contrast to clouds, the albedo from ice decreases as surface temperature rises. This means that the ice cover can alter abruptly due to the ice albedo affect. It is changes in ice cover, not ocean currents, that causes abrupt climate change, especially sea ice which is thin and even (Gildor and Tziperaman, 2003.)

Why then does Manabe and Wetherald's model work? They added convection to their model by setting the maximum lapse rate to 6.5 K/km. But convection is inevitable in an atmosphere heated at its base by solar radiation. Thus by limiting the lapse rate they were in fact setting it to 6.5 K/km. Since this is the observed mean lapse rate, then it was inevitable that their model matched reality.

It is also claimed that the current models get the OLR spectrum correct. If you investigate this using the model produced by David Archer of RealClimate, you will find a field named "Locality". This is where you select the lapse rate profile that will give you the appropriate OLR spectrum. Real Climate??? It is more like smoke and mirrors!