2020/07/01 by Andrey A. Popov, Popov, Andrey A, Changhong Mou +5 · 2 citations
Computer Science · Earth and Planetary Sciences · #62F15 #FOS: Mathematics #Meteorological Phenomena and Simulations #Numerical Analysis (math.NA) #Oceanographic and Atmospheric Processes #Optimization and Control (math.OC) #Target Tracking and Data Fusion in Sensor Networks
paper · pdf · doi:10.48550/arxiv.2007.00793
openalex publication_date 2020/07/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
This work develops a new multifidelity ensemble Kalman filter (MFEnKF)\nalgorithm based on linear control variate framework. The approach allows for\nrigorous multifidelity extensions of the EnKF, where the uncertainty in coarser\nfidelities in the hierarchy of models represent control variates for the\nuncertainty in finer fidelities. Small ensembles of high fidelity model runs\nare complemented by larger ensembles of cheaper, lower fidelity runs, to obtain\nmuch improved analyses at only small additional computational costs. We\ninvestigate the use of reduced order models as coarse fidelity control variates\nin the MFEnKF, and provide analyses to quantify the improvements over the\ntraditional ensemble Kalman filters. We apply these ideas to perform data\nassimilation with a quasi-geostrophic test problem, using direct numerical\nsimulation and a corresponding POD-Galerkin reduced order model. Numerical\nresults show that the two-fidelity MFEnKF provides better analyses than\nexisting EnKF algorithms at comparable or reduced computational costs.\n