Article
Mechanics
Zhiwen Deng, Chuangxin He, Yingzheng Liu
Summary: This paper focuses on the optimal sensor placement strategy based on a deep neural network for turbulent flow recovery within the data assimilation framework of the ensemble Kalman filter. The results demonstrate the effectiveness and robustness of the proposed strategy, showing that RANS models with EnKF augmentation were substantially improved over their original counterparts. The study concludes that the DNN-based OSP with the selection of the five most sensitive sensors can efficiently reduce the number of sensors while achieving similar or better assimilated performance.
Article
Mathematics, Applied
Alexander Wikner, Jaideep Pathak, Brian R. Hunt, Istvan Szunyogh, Michelle Girvan, Edward Ott
Summary: This study discusses the forecasting of chaotic dynamical systems using noisy partial measurements data, with a focus on combining machine learning with knowledge-based models to improve predictions. By assimilating synthetic data and training machine learning models with partial measurements, it shows potential to correct imperfections in knowledge-based models and improve forecasting accuracy.
Article
Engineering, Marine
Shaokun Deng, Zheqi Shen, Shengli Chen, Renxi Wang
Summary: The initial ensemble has an impact on the performance of ensemble-based assimilation techniques. The differences in the initial ensemble affect the convergence rate of assimilation, but all experiments eventually reach convergence. Sea surface height and sea surface salinity are more sensitive to the initial ensemble. The white-noise perturbation scheme has the largest effect, and the influence of different initial ensembles on sea surface height is concentrated in the region of the Antarctic Circumpolar Current.
JOURNAL OF MARINE SCIENCE AND ENGINEERING
(2022)
Article
Cardiac & Cardiovascular Systems
Dario De Marinis, Dominik Obrist
Summary: The proposed data assimilation methodology aims to enhance the spatial and temporal resolution of voxel-based data obtained from biomedical imaging modalities, specifically focusing on turbulent blood flow assessment in large vessels. The methodology, utilizing a Stochastic Ensemble Kalman Filter approach, combines observed flow fields with numerical simulations to improve the accuracy of flow field predictions. Validation against canonical flows and application to a clinically relevant scenario demonstrate the potential of the method to enhance 4D flow MRI data for future use.
FRONTIERS IN CARDIOVASCULAR MEDICINE
(2021)
Article
Meteorology & Atmospheric Sciences
Eviatar Bach, Michael Ghil
Summary: Data assimilation aims to optimally combine partial and noisy model forecasts and observations. Multi-model data assimilation generalizes the variational or Bayesian formulation of the Kalman filter and is proven to be the minimum variance linear unbiased estimator. In this study, a multi-model ensemble Kalman filter (MM-EnKF) based on this framework is formulated and implemented. The MM-EnKF can combine multiple model ensembles for both data assimilation and forecasting in a flow-dependent manner by providing adaptive model error estimation and matrix-valued weights for the separate models and observations. Numerical experiments using the Lorenz96 model show that the MM-EnKF results in significant error reductions compared to the best model and an unweighted multi-model ensemble in terms of probabilistic and deterministic metrics.
JOURNAL OF ADVANCES IN MODELING EARTH SYSTEMS
(2023)
Article
Astronomy & Astrophysics
Angelica M. Castillo Tibocha, Jana de Wiljes, Yuri Y. Shprits, Nikita A. Aseev
Summary: This study applies ensemble Kalman filter for data assimilation in the radiation belts, developing two new methods and validating their accuracy. The use of split-operator technique allows inclusion of more physical processes in simulations and improves computational efficiency.
SPACE WEATHER-THE INTERNATIONAL JOURNAL OF RESEARCH AND APPLICATIONS
(2021)
Article
Meteorology & Atmospheric Sciences
Troy Arcomano, Istvan Szunyogh, Alexander Wikner, Jaideep Pathak, Brian R. Hunt, Edward Ott
Summary: This paper describes the implementation of a combined hybrid-parallel prediction approach on a low-resolution atmospheric global circulation model. The hybrid model, which combines a physics-based numerical model with a machine learning component, produces more accurate forecasts for various atmospheric variables compared to the host model. Furthermore, the hybrid model exhibits smaller systematic errors and more realistic temporal variability in simulating the climate.
JOURNAL OF ADVANCES IN MODELING EARTH SYSTEMS
(2022)
Article
Meteorology & Atmospheric Sciences
Lili Lei, Yangjinxi Ge, Zhe-Min Tan, Yi Zhang, Kekuan Chu, Xin Qiu, Qifeng Qian
Summary: This study evaluates the ensemble Kalman filter (EnKF) combined with the Advanced Research Weather Research and Forecasting model (WRF) for western North Pacific typhoons in 2016. The results show that the WRF/EnKF system provides better ensemble forecasts and higher predictability for typhoon intensity compared to NCEP and ECMWF ensemble forecasts.
ADVANCES IN ATMOSPHERIC SCIENCES
(2022)
Article
Engineering, Electrical & Electronic
M. A. Gonzalez-Cagigal, J. A. Rosendo-Macias, A. Gomez-Exposito
Summary: This research presents a state estimation approach using Kalman filtering to identify the phase to which single-phase customers are connected in three-phase distribution grids. The study compares different nonlinear formulations of the Kalman filter and shows that the ensemble Kalman filter provides better estimation results as the system size increases. The accuracy, robustness, and limitations of the estimator are also tested with consideration of measurement errors.
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS
(2021)
Article
Mathematics, Applied
Theresa Lange
Summary: This paper provides a rigorous derivation of the ensemble Kalman-Bucy filter and the ensemble transform Kalman-Bucy filter in the case of nonlinear, unbounded model and observation operators. It also establishes convergence rates in terms of the discretisation step size. The paper simultaneously establishes the well-posedness and accuracy of both the continuous-time and the discrete-time filtering algorithms.
Article
Meteorology & Atmospheric Sciences
Elias D. Nino-Ruiz, Randy S. Consuegra S. Ortega, Magdalena Lucini
Summary: This paper presents the efficient and practical implementation of sequential data assimilation methods for the SPEEDY Model into climate prediction. The computational implementation of SPEEDY blends the time integrator and the spatial discretization to accelerate algebraic computations. Augmented vector states are used to propagate analysis innovations from positions to velocities.
QUARTERLY JOURNAL OF THE ROYAL METEOROLOGICAL SOCIETY
(2023)
Article
Water Resources
Andrew Pensoneault, Witold F. Krajewski, Nicolas Velasquez, Xueyu Zhu, Ricardo Mantilla
Summary: This paper discusses the application of data assimilation techniques in hydrology, focusing on the potential of EnKF and its extensions in sequential state estimation and Bayesian inverse problems. The authors improve the streamflow in a virtual catchment using the EKI algorithm and demonstrate its favorable performance.
ADVANCES IN WATER RESOURCES
(2023)
Article
Multidisciplinary Sciences
Kevin Raeder, Timothy J. Hoar, Mohamad El Gharamti, Benjamin K. Johnson, Nancy Collins, Jeffrey L. Anderson, Jeff Steward, Mick Coady
Summary: An ensemble Kalman filter reanalysis data set with a global, 80 member ensemble spanning from 2011 to 2019 is archived, providing opportunities for robust statistical analysis and machine learning training.
SCIENTIFIC REPORTS
(2021)
Article
Meteorology & Atmospheric Sciences
Jeffrey S. Whitaker, Anna Shlyaeva, Stephen G. Penny
Summary: This study compares two methods for incorporating a time-invariant, high-rank covariance estimate in an ensemble-based data assimilation system: the hybrid-covariance approach and the hybrid-gain approach. The results show that the simpler and less expensive hybrid-gain approach can achieve similar performance if the incremental normal-mode balance constraint applied to the ensemble-part of the hybrid-covariance update is turned off.
JOURNAL OF ADVANCES IN MODELING EARTH SYSTEMS
(2022)
Article
Computer Science, Interdisciplinary Applications
Gaukhar Shaimerdenova, Hakon Hoel, Raul Tempone
Summary: In this work, a highly efficient filtering method called multi-index EnKF (MIEnKF) is proposed by combining ideas from multi-index Monte Carlo and ensemble Kalman filtering. The MIEnKF method is based on independent samples of four-coupled EnKF estimators on a multi-index hierarchy of resolution levels, serving as an extension of the multilevel EnKF (MLEnKF) method developed by the same authors in 2020. Numerical verifications demonstrate that MIEnKF is more tractable and efficient than EnKF and MLEnKF.
JOURNAL OF COMPUTATIONAL PHYSICS
(2022)