Abstract

Ensemble Kalman Filter in the framework of Data Assimilation Research Testbed (DART) has been successfully implemented into a 2D kinematic flux-transport dynamo model by Dikpati and colleagues in order to do a parameter estimation, the parameter being the meridional flow-speed as function of time... Show moreEnsemble Kalman Filter in the framework of Data Assimilation Research Testbed (DART) has been successfully implemented into a 2D kinematic flux-transport dynamo model by Dikpati and colleagues in order to do a parameter estimation, the parameter being the meridional flow-speed as function of time. They performed several 'Observing System Simulation Experiments' (OSSEs), and showed that an optimal reconstruction of time-series of meridional flow-speed can be obtained by using 16 ensemble members and only one surface magnetic observation with 30% observational error. Error in reconstruction can be reduced by increasing the ensemble size and number of observations. However, this parameter reconstruction has been found to be sensitive to locations from where observational data are taken. While assimilation of low-latitudes' surface poloidal magnetic field data can produce good reconstruction, medium-frequency oscillations appear in time-series of reconstructed flow-speed if tachocline toroidal field data are assimilated. These oscillations occur primarily because tachocline toroidal fields change very little during an assimilation interval taken to be 15 days, due to changes in meridional flow. A Babcock-Leighton dynamo model's response time to changes in meridional flow-speed is a few months. We show here that rms error in reconstruction can be significantly reduced if model's response time is taken into consideration in assimilation of tachocline toroidal field data Show less