A Stochastic Energy Budget Model Using Physically Based Red Noise

TL;DR

This paper introduces a physically based stochastic energy budget model using red noise processes, emphasizing the importance of appropriate stochastic term selection for realistic meteorological modeling.

Contribution

It proposes a novel method to incorporate red noise into energy budget models, moving beyond white noise assumptions, with application to ice core data analysis.

Findings

Spectral analysis reveals red noise characteristics in ice core data.

Fitted Ornstein-Uhlenbeck processes define stochastic model parameters.

Careful stochastic term selection is crucial for physically meaningful models.

Abstract

A method to describe unresolved processes in meteorological models by physically based stochastic processes (SP) is proposed by the example of an energy budget model (EBM). Contrary to the common approach using additive white noise, a suitable variable within the model is chosen to be represented by a SP. Spectral analysis of ice core time series shows a red noise character of the underlying fluctuations. Fitting Ornstein Uhlenbeck processes to the observed spectrum defines the parameters for the stochastic dynamic model (SDM). Numerical simulations for different sets of ice core data lead to three sets of strongly differing systems. Pathwise, statistical and spectral analysis of these models show the importance of carefully choosing suitable stochastic terms in order to get a physically meaningful SDM.