The role of additive and multiplicative noise in filtering complex dynamical systems

Georg A. Gottwald, John Harlim

Research output: Contribution to journalArticle

15 Citations (Scopus)

Abstract

Covariance inflation is an ad hoc treatment that is widely used in practical real-time data assimilation algorithms to mitigate covariance underestimation owing to model errors, nonlinearity, or/and, in the context of ensemble filters, insufficient ensemble size. In this paper, we systematically derive an effective 'statistical' inflation for filtering multi-scale dynamical systems with moderate scale gap, ε = O(10-1), to the case of no scale gap with ε = O(1), in the presence of model errors through reduced dynamics from rigorous stochastic subgridscale parametrizations. We will demonstrate that for linear problems, an effective covariance inflation is achieved by a systematically derived additive noise in the forecast model, producing superior filtering skill. For nonlinear problems, we will study an analytically solvable stochastic test model, mimicking turbulent signals in regimes ranging from a turbulent energy transfer range to a dissipative range to a laminar regime. In this context, we will show that multiplicative noise naturally arises in addition to additive noise in a reduced stochastic forecast model. Subsequently, we will show that a 'statistical' inflation factor that involves mean correction in addition to covariance inflation is necessary to achieve accurate filtering in the presence of intermittent instability in both the turbulent energy transfer range and the dissipative range.

Original languageEnglish (US)
Article number20130096
JournalProceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences
Volume469
Issue number2155
DOIs
StatePublished - Jul 8 2013

Fingerprint

Complex Dynamical Systems
Multiplicative Noise
Additive Noise
Inflation
dynamical systems
Dynamical systems
Filtering
Additive noise
Model Error
Energy transfer
Energy Transfer
forecasting
Range of data
Forecast
Ensemble
energy transfer
Stochastic models
Data Assimilation
assimilation
Parametrization

All Science Journal Classification (ASJC) codes

  • Mathematics(all)
  • Engineering(all)
  • Physics and Astronomy(all)

Cite this

@article{a8dc59b7a53242a98d3acf8a331dcb02,
title = "The role of additive and multiplicative noise in filtering complex dynamical systems",
abstract = "Covariance inflation is an ad hoc treatment that is widely used in practical real-time data assimilation algorithms to mitigate covariance underestimation owing to model errors, nonlinearity, or/and, in the context of ensemble filters, insufficient ensemble size. In this paper, we systematically derive an effective 'statistical' inflation for filtering multi-scale dynamical systems with moderate scale gap, ε = O(10-1), to the case of no scale gap with ε = O(1), in the presence of model errors through reduced dynamics from rigorous stochastic subgridscale parametrizations. We will demonstrate that for linear problems, an effective covariance inflation is achieved by a systematically derived additive noise in the forecast model, producing superior filtering skill. For nonlinear problems, we will study an analytically solvable stochastic test model, mimicking turbulent signals in regimes ranging from a turbulent energy transfer range to a dissipative range to a laminar regime. In this context, we will show that multiplicative noise naturally arises in addition to additive noise in a reduced stochastic forecast model. Subsequently, we will show that a 'statistical' inflation factor that involves mean correction in addition to covariance inflation is necessary to achieve accurate filtering in the presence of intermittent instability in both the turbulent energy transfer range and the dissipative range.",
author = "Gottwald, {Georg A.} and John Harlim",
year = "2013",
month = "7",
day = "8",
doi = "10.1098/rspa.2013.0096",
language = "English (US)",
volume = "469",
journal = "Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences",
issn = "0080-4630",
publisher = "Royal Society of London",
number = "2155",

}

The role of additive and multiplicative noise in filtering complex dynamical systems. / Gottwald, Georg A.; Harlim, John.

In: Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, Vol. 469, No. 2155, 20130096, 08.07.2013.

Research output: Contribution to journalArticle

TY - JOUR

T1 - The role of additive and multiplicative noise in filtering complex dynamical systems

AU - Gottwald, Georg A.

AU - Harlim, John

PY - 2013/7/8

Y1 - 2013/7/8

N2 - Covariance inflation is an ad hoc treatment that is widely used in practical real-time data assimilation algorithms to mitigate covariance underestimation owing to model errors, nonlinearity, or/and, in the context of ensemble filters, insufficient ensemble size. In this paper, we systematically derive an effective 'statistical' inflation for filtering multi-scale dynamical systems with moderate scale gap, ε = O(10-1), to the case of no scale gap with ε = O(1), in the presence of model errors through reduced dynamics from rigorous stochastic subgridscale parametrizations. We will demonstrate that for linear problems, an effective covariance inflation is achieved by a systematically derived additive noise in the forecast model, producing superior filtering skill. For nonlinear problems, we will study an analytically solvable stochastic test model, mimicking turbulent signals in regimes ranging from a turbulent energy transfer range to a dissipative range to a laminar regime. In this context, we will show that multiplicative noise naturally arises in addition to additive noise in a reduced stochastic forecast model. Subsequently, we will show that a 'statistical' inflation factor that involves mean correction in addition to covariance inflation is necessary to achieve accurate filtering in the presence of intermittent instability in both the turbulent energy transfer range and the dissipative range.

AB - Covariance inflation is an ad hoc treatment that is widely used in practical real-time data assimilation algorithms to mitigate covariance underestimation owing to model errors, nonlinearity, or/and, in the context of ensemble filters, insufficient ensemble size. In this paper, we systematically derive an effective 'statistical' inflation for filtering multi-scale dynamical systems with moderate scale gap, ε = O(10-1), to the case of no scale gap with ε = O(1), in the presence of model errors through reduced dynamics from rigorous stochastic subgridscale parametrizations. We will demonstrate that for linear problems, an effective covariance inflation is achieved by a systematically derived additive noise in the forecast model, producing superior filtering skill. For nonlinear problems, we will study an analytically solvable stochastic test model, mimicking turbulent signals in regimes ranging from a turbulent energy transfer range to a dissipative range to a laminar regime. In this context, we will show that multiplicative noise naturally arises in addition to additive noise in a reduced stochastic forecast model. Subsequently, we will show that a 'statistical' inflation factor that involves mean correction in addition to covariance inflation is necessary to achieve accurate filtering in the presence of intermittent instability in both the turbulent energy transfer range and the dissipative range.

UR - http://www.scopus.com/inward/record.url?scp=84878968726&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84878968726&partnerID=8YFLogxK

U2 - 10.1098/rspa.2013.0096

DO - 10.1098/rspa.2013.0096

M3 - Article

AN - SCOPUS:84878968726

VL - 469

JO - Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences

JF - Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences

SN - 0080-4630

IS - 2155

M1 - 20130096

ER -