DI31B-4281:
Assimilation of the Observed SV with Geodynamo Modeling and Sensitivities of the Core Flow Beneath the Cmb

Wednesday, 17 December 2014
Weijia Kuang, GSFC / NASA, Greenbelt, MD, United States and Andrew Tangborn, Univ Maryland-Baltimore County, Greenbelt, MD, United States
Abstract:
One challenge in geomagnetic data assimilation is that the observational records are relatively short: they are approximately 6 times to the typical time scales of non-dipole geomagnetic field, and approximately 10 times to that of the dipole field. However, surface geomagnetic observations are made at time scales at least three orders of magnitude shorter than those of the geomagnetic secular variation. Therefore, the field and its secular variation (SV), i.e. the rate of changes in time, are “concurrently” measured, allowing us to assimilate both pieces of information into geodynamo modeling. This can be achieved via construction of a new observation operator for the velocity field beneath the CMB, and modification of the existing observation operator for the magnetic field.

Determination of the observation operator for the velocity field is very complex, as the constraints from the observed SV are two orders of magnitude less than the velocity field state variables in dynamo simulation. Therefore, it depends on appropriate understanding of the responses of the core flow to the observed SV.

For this purpose, we carried out several assimilation experiments with and without assimilation of SV. In particular, the observed SV is only utilized for correcting the model forecasts in the magnetic field, leaving the velocity field forecast unconstrained. Our experiment shows that, the large-scale axisymmetric toroidal velocity responds the strongest to the observed SV, following by a suite of small-scale velocity components (with the span from spherical harmonic degree 10 to 20). The non-axisymmetric part of the poloidal flow is also very sensitive to the SV constraints. However, such responses are more evenly spread out from large to small-scale flow patterns. These assimilation results will be used for construction of the velocity field observational operator, and can be also applied to core-flow inversion studies in general.