Data Assimilation of Rayleigh-Bénard Convection using Thermal Measurements
ORAL
Abstract
Nudging1 is a continuous data assimilation technique which introduces a penalty term in the equations of motion to keep the evolution of the flow close to a given set of measurements. We employ a nudging2 scheme to reconstruct the temperature and velocity fields in Rayleigh-Bénard convection using only thermal data.
We show that for moderate Rayleigh number, the nudged flow is able to fully synchronize to the reference flow and accurately reconstruct both the temperature and velocity; In more turbulent cases with higher Rayleigh number, where the underlying correlations between the velocity and temperature are diminished, the nudged flow synchronizes only partially, as only the temperature field is accurately reconstructed. We also assess how the nudging fares when the data becomes sparse, allowing us to determine the minimum spacing between measurements required to fully synchronize the flows.
1Gesho, Masakazu, Eric Olson, and Edriss S. Titi. Commun. Comput. Phys. 19, no. 4 (2016): 1094-1110.
2 Clark Di Leoni, Patricio, Andrea Mazzino, and Luca Biferale. Phys. Rev. X 10, no. 1 (2020): 011023.
We show that for moderate Rayleigh number, the nudged flow is able to fully synchronize to the reference flow and accurately reconstruct both the temperature and velocity; In more turbulent cases with higher Rayleigh number, where the underlying correlations between the velocity and temperature are diminished, the nudged flow synchronizes only partially, as only the temperature field is accurately reconstructed. We also assess how the nudging fares when the data becomes sparse, allowing us to determine the minimum spacing between measurements required to fully synchronize the flows.
1Gesho, Masakazu, Eric Olson, and Edriss S. Titi. Commun. Comput. Phys. 19, no. 4 (2016): 1094-1110.
2 Clark Di Leoni, Patricio, Andrea Mazzino, and Luca Biferale. Phys. Rev. X 10, no. 1 (2020): 011023.
–
Presenters
-
Lokahith Narendra Agasthya
INFN-Rome
Authors
-
Lokahith Narendra Agasthya
INFN-Rome
-
Patricio Clark Di Leoni
Johns Hopkins University, Department of Mechanical Engineering, Johns Hopkins University, Baltimore, Maryland 21218, USA, Dept. of Mechanical Engineering, Johns Hopkins University, Baltimore, Maryland 21218, USA
-
Luca Biferale
University of Rome "Tor Vergata", Italy, University of Rome "Tor Vergata", INFN, University of Rome Tor Vergata, INFN - Rome, Department of Physics & INFN, University of Rome Tor Vergata, Via della Ricerca Scientifica 1, 00133 Rome, Italy