HR: 1340h
AN: GP33A-0927    [Abstracts]
TI: Application of Ensemble Techniques in Geomagnetic Data Assimilation
AU: * Sun, Z
EM: sunzhib1@umbc.edu
AF: Joint Center for Earth Systems Technology at University of Maryland-Baltimore County, 1000 Hilltop Circle, Baltimore, MD 21250, United States
AU: Tangborn, A
EM: tangborn@gmao.gsfc.nasa.gov
AF: Joint Center for Earth Systems Technology at University of Maryland-Baltimore County, 1000 Hilltop Circle, Baltimore, MD 21250, United States
AU: Kuang, W
EM: Weijia.Kuang-1@nasa.gov
AF: Planetary Geodynamics Branch-GSFC, Goddard Space Flight Center, Code 698, Greenbelt, MD 20771, United States
AB: Geomagnetic data assimilation is a recent application of data assimilation in which an improved estimate of the state of the Earth's core is achieved by combing surface geomagnetic observations with a geodynamo model. Provided that good estimates of forecast model and observation error statistics are available, an optimal estimate can potentially be obtained. Geomagnetic field observations have well understood error characteristics. On the other hand, we have begun to develop the methods to estimate the error statistics of the MoSST core dynamics model, a geodynamo model that uses spherical harmonics and finite differences for spatial derivative approximation. Together with a limited part of the poloidal magnetic field observed at the Earth's surface, we need to apply the estimated error statistics of MoSST core dynamics model correctly into a data assimilation system. We have developed a geomagnetic data assimilation system, in which the forecast error covariances are estimated using an ensemble of model solutions. By analyzing the covariances, we know not only how deep the poloidal magnetic field inside the core should be corrected by the surface observations, but also how other state variables, i.e. the remaining poloidal field, the toroidal magnetic field, the velocity field and the density perturbation should be corrected. We use an ensemble method to estimate the forecast error covariance by perturbing an ensemble of initial states. Choosing an appropriate perturbation for the model runs is a critical part of ensemble methods. In particular, perturbations which do not satisfy all of the boundary conditions, or which do not satisfy the conservation equations for all of the variables, may introduce spurious oscillations or spikes in the solutions. In the present work we investigate a number of alternate perturbation strategies for the assimilation system, and compare their impact on the assimilation system by a series of Observing System Simulation Experiments (OSSE's).
DE: 3315 Data assimilation
DE: 4400 NONLINEAR GEOPHYSICS (3200, 6944, 7839)
DE: 5440 Magnetic fields and magnetism
SC: Geomagnetism and Paleomagnetism [GP]
MN: 2007 Fall Meeting