Why not combine both dynamical and statistical methodologies?
These methods are complementary, and we believe current assessments of the state of the science relating hurricanes to global warming need to take both sources of information into account. We need to learn how to best optimize the information coming from both approaches. In particular, dynamical models can help us choose among competing statistical models exhibiting similar skill over the observed record, but built using different sets of predictors. For example, while impressive fits to interannual variations in storm activity in the Atlantic can be obtained using local tropical Atlantic sea surface temperatures (SSTs) as a predictor, Emanuel (2005), equally impressive fits can be obtained using as a predictor the difference between the tropical Atlantic SST and the tropical mean SST — see Swanson (2008). Models intermediate in structure between the simplest statistical approaches and fully dynamical approaches are also possible. The recent work of Emanuel (2008) is a good example.
