Ensemble Kalman filter analysis of magnetic field models during the CHAMP-Swarm gap

Abstract

Between the de-orbiting of CHAMP in September 2010 and the launch of Swarm in November 2013, there was a lack of satellite vector magnetic field data to use for main field modelling. During this period the difference between field models derived at the time and retrospective analysis (using data both before and after the vector gap) rose to around 20 nT root-mean-square (RMS). We use Ensemble Kalman Filtering (EnKF) to combine models of steady flow at the outer core surface with magnetic field models derived from the period when no vector satellite data were available. Since we find that the field models produced during periods without vector satellite data are just as good as the annual predictions from a flow model, there appears, at present, to be no overall benefit to using EnKF to improve field forecasting. This will remain the case until flow modelling can better forecast secular variation.