Weather noise forcing of surface climate variability

A model-based method to evaluate the role of weather noise forcing of low-frequency variability of surface properties, including SST, surface currents, land surface temperature, and soil moisture, is presented. In this procedure, an interactive ensemble coupled model, in which the ocean and land components are coupled to the mean of an ensemble of atmospheric models, is externally forced by the time-dependent surface fluxes due to the weather noise. The coupling to the ensemble mean atmosphere has the effect of strongly reducing the internally generated atmospheric weather noise so that, for example, observed weather noise forcing can meaningfully be prescribed. The surface flux due to weather noise is determined by removing the ensemble mean surface flux response of an ensemble of atmospheric models to specified surface boundary conditions from the total surface flux. When the system is entirely noise forced and the forcing is specified in this manner, the interactive ensemble model will reproduce the observed evolution of the surface properties. The method is illustrated in the context of two models: 1) a well-known simple coupled model of low-frequency climate variability, which considers the response of a slab thermodynamic surface model coupled to a thermodynamic atmosphere, and 2) a coupled GCM. The simple model is applied to the diagnosis of both observational and model-generated data, while the coupled GCM example illustrates the potential of the method if both model and data are perfect.