DAFI - Data Assimilation and Field Inversion

Ensemble based data-assimilation and field inversion.

Developed at Dr. Heng Xiao's group at Virginia Tech. In December 2022, Dr. Xiao moved to University of Stuttgart to become Chair of Data-Driven Fluid Dynamics

Website: https://dafi.readthedocs.io

Cite: C. A. Michelén Ströfer, X-L. Zhang, H. Xiao. DAFI: An open-source framework for ensemble-based data assimilation and field inversion. Communications in Computational Physics 29, pp. 1583-1622, 2021. DOI: 10.4208/cicp.OA-2020-0178. Also available at: arxiv: 2012.02651.

List of publications using DAFI:

• X.-L. Zhang, H. Xiao, X. Luo, G. He. Ensemble Kalman method for learning turbulence models from indirect observation data. Journal of Fluid Mechanics, 949(A26), 2022.

• C. A. Michelén Ströfer, X-L. Zhang, H. Xiao, O. Coutier-Delgosha. Enforcing boundary conditions on physical fields in Bayesian inversion. Computer Methods in Applied Mechanics and Engineering 367, 113097, 2020. DOI: 10.1016/j.cma.2020.113097. Also available at: arxiv: 1911.06683.

• X.-L. Zhang, C. A. Michelén Ströfer, H. Xiao. Regularization of ensemble Kalman methods for inverse problems. Journal of Computational Physics, 416, 109517, 2020. DOI: 10.1016/j.jcp.2020.109517. Also available at: arxiv: 1910.01292.

• X.-L. Zhang, H. Xiao, T. Gomez, O. Coutier-Delgosha. Evaluation of ensemble methods for quantifying uncertainties in steady-state CFD applications with small ensemble sizes. Computers & Fluids, 203, 104530, 2020. DOI: 10.1016/j.compfluid.2020.104530. Also available at: arxiv: 2004.05541.

• X.-L. Zhang, H. Xiao, G. He, S. Wang. Assimilation of disparate data for enhanced reconstruction of turbulent mean flows. Computers & Fluids, 224, 104962, 2021. DOI: 10.1016/j.compfluid.2021.104962.

• X.-L. Zhang, H. Xiao, G. He. Assessment of Regularized Ensemble Kalman Method for Inversion of Turbulence Quantity Fields. AIAA Journal, In Press, 2021. DOI: 10.2514/1.J060976.

• X.-L. Zhang, H. Xiao, T. Wu, G. He. Acoustic Inversion for Uncertainty Reduction in Reynolds-Averaged Navier–Stokes-Based Jet Noise Prediction. AIAA Journal, In Press, 2021. DOI: 10.2514/1.J060876.

Contributors:

• Carlos A. Michelén Ströfer
• Xinlei Zhang
• Jianxun Wang
• Rui Sun
• Jinlong Wu

Contact: Carlos A. Michelén Ströfer; Heng Xiao