Report from breakout group 1: How can work in nonlinear PDEs most benefit climate prediction?

• Main Authors: Paul Williams, Marek Wlasak
• Other Authors: The Pennsylvania State University CiteSeerX Archives
• Format: Text
• Language: English
• Published: 2009
• Subjects: Arctic; North Atlantic; North Atlantic oscillation
• Online Access: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.674.9699; http://www.met.reading.ac.uk/%7Ewilliams/publications/UKMO2009_BreakoutSession1.pdf

Andy Majda discussed a methodology for forecasting low frequency teleconnection patterns, such as the Arctic Oscillation (AO) and North Atlantic Oscillation (NAO). By applying a fluctuation-dissipation theorem (FDT), the climate response to small external forcing can be estimated from well-chosen statistics of the present climate. FDTs centre on identifying the appropriate response function, using an ergodic assumption and assuming that the evolution is in statistical equilibrium. The technique may be generalised to calculate accurate variance and mean estimates. Sensitivity information can also be obtained such that the forcing can be readily determined from a given response. We may also use FDTs to infer the regional perturbations that contain the most information. FDTs require strong mixing. They have been shown to work in the Lorenz (1996) model with 40 modes and in other simple climate models. They are also expected to work in more complex climate models. The validity of the response function is limited by the fidelity of the climate model from which it is generated. Climate models are numerical discretisations with finite