Empirical Modeling and Stochastic Simulation of Sea-Level Pressure Variability

Abstract:

The scope of this work is stochastic emulation of surrogate sea-level pressure (SLP) for the use in error estimation and statistical prediction studies. The input SLP data set whose statistics is to be emulated was taken from the 1979–2013 ERA Interim Reanalysis at full 6-hourly temporal and 0.75º spatial resolutions over the Northern Hemisphere. Upon subtracting seasonal climatology, the SLP anomalies (SLPA) were projected onto the subspace of 1000 leading empirical orthogonal functions (EOFs) of the daily-mean SLPA, which account for the vast majority (>99%) of the full 6-hourly fields’ variance for each season. The main step of our methodology is the estimation of a linear autoregressive moving average (ARMA) empirical model for the daily SLPA principal components (PCs) via regularized multiple linear regression (MLR); this model was driven, at the stage of simulation, by state-dependent (multiplicative) noise. Finally, we have also developed and implemented a diagnostic statistical scheme for accurate interpolation of simulated daily SLPA anomalies to 6-hourly temporal resolution. Upon transforming the simulated 6-hourly SLPA PCs into the physical space and adding seasonal climatology, the resulting SLP variability was compared to the actual variability in the ERA Interim Reanalysis. We show that our empirical model produces independent realizations of SLP variability, which are nearly indistinguishable from the observed variability over a wide range of statistical measures; these measures include, among other things, spatial patterns of band-pass and low-pass filtered variability, as well as diverse characteristics of mid-latitude cyclone tracks.