Watts Up With That is looking into complexity and finds prediction in complex situations difficult according to the IPCC.

Third Assessment Report: “In sum, a strategy must recognise what is possible. In climate research and modeling, we should recognize that we are dealing with a coupled nonlinear chaotic system, and therefore that long-term prediction of future climate states is not possible.”

It all starts from the simple idea that increasing internal radiation reflection (Greenhouse Gas Theory) increases temperature. Fine so far. But there are complications. The climate system is not simple physics. For instance: surface water absorbs energy. And the Earth’s surface has a LOT of water. Well the water evaporates and you get clouds. Clouds are very complicated. Sometimes the water vapor precipitates out of the atmosphere and you get rain, or snow, or sleet, or hail. This is part of an energy transport system (heat pipe)in the atmosphere. Clouds also complicate the radiation picture. They reflect from both sides. Which matters day AND night on Earth but only day for incoming radiation. And that is just one aspect of the system. Vegetation varies radiation depending on type and amount.

The simple and not so simple physics of a number of climate parameters, are programmed into the climate computer models. Many of these parameters, it is acknowledged, are not completely understood or that there is serious contentious debate about in the scientific literature. ie aerosols, clouds, solar pacific and atlantic oscillations, volcanoes, etc,etc
Engineers (or economists now, perhaps) will advice climate scientists, model are not reality, reality is often more complicated than any computer model. Take a step back, view with hindsight with respect to risk in the financial markets. At the trouble the cream of the last few decades of science graduates – turned computer modellers – left the world’s economy in, following the modelling of credit risk amongst many other economic assumptions.

Engineers have always been the biggest sceptics (I prefer the Brit spelling) of the CO2 causes global warming hypothesis.
Engineers spend decades in efforts to match simple deterministic systems to complex environments.
Back in the 80s I (electronics engineer) used to worry about second order effects (deviations from simple laws) caused by the non-linearity of materials. We are now in third order territory with occasional forays into fourth order effects. Climate is like 14th or 40th order stuff. And very non-linear. It is possible (not likely) for 12th order effects to have first order results (chaos).
Is it possible to do decent predictions in short time frames? Maybe. I saw Piers Corbin on Nightline the other day and he has claimed to have predicted much recent weather and predicts a cold winter in the Northern Hemisphere.

Piers Corbyn’s presentation showed the major advances in power and skill now achieved by his Solar-Lunar-Action-Technique (SLAT) of weather & climate forecasting which now includes the ability to predict from months ahead extreme events all over the world and changes in the Jet Stream such as those which caused the West Russian heatwave and the Pakistan super-deluges and floods and marked their ending in mid August 2010. In his presentation Piers showed a film of the double sunspot superfast solar coronal ejection on 14th August and the consequent Sudden Ionospheric Disturbance and the predicted jet stream disruption – See:
For WeatherAction summer forecasts 2007, 08, 09 and winter forecasts 2008/9 and 09/10 which beat all-comers see:

He seems to be doing OK so far. BTW he is a “CO2 causes global warming” sceptic.

Some urls:
And some books:
Red Hot Lies: How Global Warming Alarmists Use Threats, Fraud, and Deception to Keep You Misinformed
The Chilling Stars: A New Theory of Climate Change
Heaven and Earth: Global Warming, the Missing Science
Cross Posted at Power and Control