Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Data assimilation empowered neural network parametrizations for subgrid processes in geophysical flows

Journal Article · · Physical Review Fluids (Online)
In the past couple of years, there has been a proliferation in the use of machine learning approaches to represent subgrid-scale processes in geophysical flows with an aim to improve the forecasting capability and to accelerate numerical simulations of these flows. Despite its success for different types of flow, the online deployment of a data-driven closure model can cause instabilities and biases in modeling the overall effect of subgrid-scale processes, which in turn leads to inaccurate prediction. To tackle this issue, we exploit the data assimilation technique to correct the physics-based model coupled with the neural network as a surrogate for unresolved flow dynamics in multiscale systems. In particular, we use a set of neural network architectures to learn the correlation between resolved flow variables and the parametrizations of unresolved flow dynamics and formulate a data assimilation approach to correct the hybrid model during their online deployment. We illustrate our framework in a set of applications of the multiscale Lorenz 96 system for which the parametrization model for unresolved scales is exactly known, and the two-dimensional Kraichnan turbulence system for which the parametrization model for unresolved scales is not known a priori. Our analysis, therefore, comprises a predictive dynamical core empowered by (i) a data-driven closure model for subgrid-scale processes, (ii) a data assimilation approach for forecast error correction, and (iii) both data-driven closure and data assimilation procedures. We show significant improvement in the long-term prediction of the underlying chaotic dynamics with our framework compared to using only neural network parametrizations for future prediction. Moreover, we demonstrate that these data-driven parametrization models can handle the non-Gaussian statistics of subgrid-scale processes, and effectively improve the accuracy of outer data assimilation workflow loops in a modular nonintrusive way.
Research Organization:
Oklahoma State University, Stillwater, OK (United States)
Sponsoring Organization:
USDOE Office of Science (SC)
Grant/Contract Number:
SC0019290
OSTI ID:
1853026
Journal Information:
Physical Review Fluids (Online), Journal Name: Physical Review Fluids (Online) Journal Issue: 5 Vol. 6; ISSN 2469-990X
Publisher:
American Physical Society (APS)Copyright Statement
Country of Publication:
United States
Language:
English

References (119)

Predicting the Future book January 2013
Deep neural networks for data-driven LES closure models journal December 2019
Backpropagation algorithms and Reservoir Computing in Recurrent Neural Networks for the forecasting of complex spatiotemporal dynamics journal June 2020
Neural networks based subgrid scale modeling in large eddy simulations journal January 2003
Reynolds-averaged Navier–Stokes equations with explicit data-driven Reynolds stress closure can be ill-conditioned journal April 2019
Turbulence model reduction by deep learning journal June 2020
Perspective on machine learning for advancing fluid mechanics journal October 2019
Spatio-temporal interpolation of altimeter-derived SSH fields using analog data assimilation: A case-study in the south china sea conference July 2017
The role of large-scale spatial patterns in the chaotic amplification of perturbations in a Lorenz’96 model journal January 2011
Subgrid-Scale Parameterization with Conditional Markov Chains journal August 2008
Maximum Likelihood Ensemble Filter: Theoretical Aspects journal June 2005
Application of Deep Convolutional Neural Networks for Detecting Extreme Weather in Climate Datasets preprint January 2016
A proof of concept for scale-adaptive parameterizations: the case of the Lorenz '96 model text January 2016
Deep Learning for Precipitation Nowcasting: A Benchmark and A New Model preprint January 2017
Super-resolution reconstruction of turbulent flows with machine learning text January 2018
Shallow Neural Networks for Fluid Flow Reconstruction with Limited Sensors text January 2019
Data-driven modelling of nonlinear spatio-temporal fluid flows using a deep convolutional generative adversarial network text January 2020
Recurrent Neural Network Architecture Search for Geophysical Emulation preprint January 2020
Optimal State Estimation book January 2006
Middle atmosphere dynamical sources of the semiannual oscillation in the thermosphere and ionosphere: MIDDLE ATMOSPHERE SOURCES OF THE T-I SAO journal January 2017
A proof of concept for scale-adaptive parametrizations: the case of the Lorenz '96 model: Proof for Scale-adaptive Parametrizations journal December 2017
A nonlinear dynamical perspective on model error: A proposal for non-local stochastic-dynamic parametrization in weather and climate prediction models journal January 2001
Computational Design for Long-Term Numerical Integration of the Equations of Fluid Motion: Two-Dimensional Incompressible Flow. Part I journal August 1997
Data Assimilation book January 2009
Structure functions in two-dimensional turbulence journal February 1992
A priori analysis on deep learning of subgrid-scale parameterizations for Kraichnan turbulence journal January 2020
Ensemble Kalman Filter for Assimilating Experimental Data into Large-Eddy Simulations of Turbulent Flows journal December 2019
Harmonizing models and observations: Data assimilation in Earth system science journal June 2020
Cloud parameterization for climate modeling: Status and prospects journal October 1989
Neural networks based subgrid scale modeling in large eddy simulations journal January 2003
Continuous data assimilation reduced order models of fluid flow journal December 2019
Data-driven modelling of nonlinear spatio-temporal fluid flows using a deep convolutional generative adversarial network journal June 2020
High-order methods for decaying two-dimensional homogeneous isotropic turbulence journal June 2012
A stable and scale-aware dynamic modeling framework for subgrid-scale parameterizations of two-dimensional turbulence journal November 2017
Parameterised non-intrusive reduced order methods for ensemble Kalman filter data assimilation journal November 2018
A turbulent eddy-viscosity surrogate modeling framework for Reynolds-averaged Navier-Stokes simulations journal September 2021
Reconstruction of unsteady viscous flows using data assimilation schemes journal July 2016
Hidden physics models: Machine learning of nonlinear partial differential equations journal March 2018
Optimal reduced space for Variational Data Assimilation journal February 2019
Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations journal February 2019
Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders journal November 2019
Enforcing statistical constraints in generative adversarial networks for modeling chaotic dynamical systems journal April 2020
A deep-learning-based surrogate model for data assimilation in dynamic subsurface flow problems journal July 2020
Combining data assimilation and machine learning to emulate a dynamical model from sparse and noisy observations: A case study with the Lorenz 96 model journal July 2020
Towards a mesoscale eddy closure journal January 2008
Approximate deconvolution large eddy simulation of a stratified two-layer quasigeostrophic ocean model journal March 2013
Sampling the posterior: An approach to non-Gaussian data assimilation journal June 2007
Filter accuracy for the Lorenz 96 model: Fixed versus adaptive observation operators journal June 2016
Lift & Learn: Physics-informed machine learning for large-scale nonlinear dynamical systems journal May 2020
Statistical mechanics of two-dimensional and geophysical flows journal June 2012
Dynamic Data Assimilation book January 2006
State estimation in wall-bounded flow systems. Part 3. The ensemble Kalman filter journal August 2011
Reynolds averaged turbulence modelling using deep neural networks with embedded invariance journal October 2016
Subgrid modelling for two-dimensional turbulence using neural networks journal November 2018
Super-resolution reconstruction of turbulent flows with machine learning journal May 2019
Prediction of turbulent heat transfer using convolutional neural networks journal November 2019
Could Machine Learning Break the Convection Parameterization Deadlock? journal June 2018
Prognostic Validation of a Neural Network Unified Physics Parameterization journal June 2018
Spatially Extended Tests of a Neural Network Parametrization Trained by Coarse‐Graining journal August 2019
A Machine Learning‐Based Global Atmospheric Forecast Model journal May 2020
Deep learning journal May 2015
Deep learning and process understanding for data-driven Earth system science journal February 2019
Inertial Ranges in Two-Dimensional Turbulence journal January 1967
Extensive chaos in the Lorenz-96 model journal December 2010
A deep learning enabler for nonintrusive reduced order modeling of fluid flows journal August 2019
A dynamic subgrid‐scale eddy viscosity model journal July 1991
A proposed modification of the Germano subgrid‐scale closure method journal March 1992
Combining machine learning with knowledge-based modeling for scalable forecasting and subgrid-scale closure of large, complex, spatiotemporal systems journal May 2020
Deep learning to represent subgrid processes in climate models journal September 2018
Earth observation satellite sensors for biodiversity monitoring: potentials and bottlenecks journal September 2014
Deep learning for physical processes: incorporating prior scientific knowledge journal December 2019
Stochastic parameterization for large eddy simulation of geophysical flows journal April 2007
Total variation diminishing Runge-Kutta schemes journal January 1998
Shallow neural networks for fluid flow reconstruction with limited sensors
  • Erichson, N. Benjamin; Mathelin, Lionel; Yao, Zhewei
  • Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, Vol. 476, Issue 2238 https://doi.org/10.1098/rspa.2020.0097
journal June 2020
Subgrid modelling for geophysical flows
  • Frederiksen, Jorgen S.; O'Kane, Terence J.; Zidikheri, Meelis J.
  • Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, Vol. 371, Issue 1982 https://doi.org/10.1098/rsta.2012.0166
journal January 2013
Assessing the scales in numerical weather and climate predictions: will exascale be the rescue?
  • Neumann, Philipp; Düben, Peter; Adamidis, Panagiotis
  • Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, Vol. 377, Issue 2142 https://doi.org/10.1098/rsta.2018.0148
journal February 2019
Nonintrusive reduced order modeling framework for quasigeostrophic turbulence journal November 2019
Turbulence model reduction by deep learning journal June 2020
Artificial neural network approach to large-eddy simulation of compressible isotropic turbulence journal May 2019
Physics-informed machine learning approach for reconstructing Reynolds stress modeling discrepancies based on DNS data journal March 2017
Searching for turbulence models by artificial neural network journal May 2017
Predictive large-eddy-simulation wall modeling via physics-informed neural networks journal March 2019
Predictions of turbulent shear flows using deep neural networks journal May 2019
Perspective on machine learning for advancing fluid mechanics journal October 2019
Modeling subgrid-scale forces by spatial artificial neural networks in large eddy simulation of turbulence journal May 2020
Model-Free Prediction of Large Spatiotemporally Chaotic Systems from Data: A Reservoir Computing Approach journal January 2018
Gradient-based learning applied to document recognition journal January 1998
Theory-Guided Data Science for Climate Change journal November 2014
Theory-Guided Data Science: A New Paradigm for Scientific Discovery from Data journal October 2017
An adaptive covariance inflation error correction algorithm for ensemble filters journal January 2007
A deterministic formulation of the ensemble Kalman filter: an alternative to ensemble square root filters journal January 2008
The maximum likelihood ensemble filter performances in chaotic systems journal January 2008
Assessment and Propagation of Model Uncertainty journal January 1995
A New Approach to Linear Filtering and Prediction Problems journal March 1960
Symmetries in the Lorenz-96 Model journal January 2019
Turbulence Modeling in the Age of Data journal January 2019
Machine Learning for Fluid Mechanics journal September 2019
Two-Dimensional Turbulence journal January 2012
Atmospheric Predictability and Two-Dimensional Turbulence journal March 1971
General Circulation Experiments with the Primitive Equations: i. the Basic Experiment* journal March 1963
Analysis Scheme in the Ensemble Kalman Filter journal June 1998
A Monte Carlo Implementation of the Nonlinear Filtering Problem to Produce Ensemble Assimilations and Forecasts journal December 1999
Subgrid-Scale Parameterization with Conditional Markov Chains journal August 2008
Moisture Vertical Structure, Column Water Vapor, and Tropical Deep Convection journal June 2009
Accelerating Progress in Global Atmospheric Model Development through Improved Parameterizations: Challenges, Opportunities, and Strategies journal July 2010
A Non-Gaussian Ensemble Filter Update for Data Assimilation journal November 2010
An Improved Strategy for the Evaluation of Cloud Parameterizations in GCMS journal October 2003
Dynamical Subgrid-Scale Parameterizations from Direct Numerical Simulations journal November 2006
Uncertainty in Model Climate Sensitivity Traced to Representations of Cumulus Precipitation Microphysics journal January 2016
Review of the Ensemble Kalman Filter for Atmospheric Data Assimilation journal December 2016
The Analog Data Assimilation journal October 2017
Maximum Likelihood Ensemble Filter: Theoretical Aspects journal June 2005
Effects of stochastic parametrizations in the Lorenz '96 system journal January 2005
Data-assisted reduced-order modeling of extreme events in complex dynamical systems journal May 2018
Ensemble-Based State Estimator for Aerodynamic Flows journal July 2018
Theory-Guided Machine Learning in Materials Science journal June 2016
A Maximum Likelihood Ensemble Filter via a Modified Cholesky Decomposition for Non-Gaussian Data Assimilation journal February 2020
Bayesian inference of chaotic dynamics by merging data assimilation, machine learning and expectation-maximization journal January 2020
Coupled online learning as a way to tackle instabilities and biases in neural network parameterizations: general algorithms and Lorenz 96 case study (v1.0) journal January 2020

Similar Records

A priori analysis on deep learning of subgrid-scale parameterizations for Kraichnan turbulence
Journal Article · Fri Jan 03 19:00:00 EST 2020 · Theoretical and Computational Fluid Dynamics · OSTI ID:1737499

Frame invariant neural network closures for Kraichnan turbulence
Journal Article · Wed Nov 16 19:00:00 EST 2022 · Physica. A · OSTI ID:2421106

Subgrid modelling for two-dimensional turbulence using neural networks
Journal Article · Wed Jan 09 19:00:00 EST 2019 · Journal of Fluid Mechanics · OSTI ID:1593570