A linear model of low-frequency variability

An alternate paradigm of low-frequency variability is investigated: variability in the extratropics is predominantly the result of low-frequency tropical diabatic heating which forces extratropical anomalies that can be additionally modified by transient eddy feedbacks. The zonally-asymmetric time-mean basic state is necessary to organize the variability. However, its importance as an energy source/sink (as stressed by Simmons et al.) is secondary. To test this hypothesis, we use a linear inverse model (LIM) which was constructed from observations of the past thirty years. Notably, this empirical-dynamical model includes tropical diabatic heating as an evolving model variable rather than as an externally specified forcing, and also includes, in effect, the feedback of extratropical weather systems on the more slowly varying circulation.

The skill of the LIM (with 37 degrees of freedom) at forecasting week 2 anomalies is competitive with that of the nonlinear medium-range forecast (MRF) model with (10^6) degrees of freedom in use at the National Centers for Environmental Prediction. This suggests that the LIM can also be used to diagnose the statistics of observed low-frequency variability. And in fact, the LIM is able to reproduce the observed lagged covariability statistics of low-frequency variability remarkably well. This quantity is badly simulated by both barotropic and baroclinic models linearized about the time-mean flow. Diabatic heating is critical to this correct temporal simulation. The spatial structure of the stochastic forcing necessary to maintain variability is discussed and compared to that required in simpler models. The energy balance of the variability is analyzed to quantify the relative importance of the different mechanisms which contribute to variability.