High-resolution empirical geomagnetic field model TS07D: Present status and outlook

Increasing sophistication of space weather modeling and forecasting tools requires further increase of spatial and temporal resolution of the empirical geomagnetic field models. A key limitation of the latter models, arising from their pre-defined modular structure, has been overcome in the model TS07D (http://geomag_field.jhuapl.edu/model/). The new model employs for the first time the expansion of the magnetic field of equatorial currents into a series of basis functions without any a priori assumptions on their evolution in time. The latter is reconstructed by fitting the instantaneous set of the model parameters with a subset of the whole database of the available magnetic field data, so-called nearest neighbors (NN), when the global state of the magnetosphere was close to the considered one in the space of the global parameters. These include the average solar wind electric field parameter vBz, Sym-H index, and its time derivative. The model also explicitly takes into account variations of the tilt angle and the solar wind ram pressure. TS07D has been successfully validated for a number of in-sample modeling cases, when the considered event (a magnetic storm) was also a part of the NN subset. Further verification of the model and the development of its forecasting versions require also out-of-sample validation tests. We present some of such tests, including THEMIS observations of the March 2008 storm. Comparison between the model and observations made by five THEMIS probes shows excellent agreement on storm scales. Moreover, every deviation arising during a substorm reveals characteristic signatures of the tail current sheet thinning and dipolarization, suggesting that TS07D can be used as a base-line model for investigations of storm-time substorms. Further development of the model, including the improvement of its field-aligned component, fully-forecasting modes, and the use of the global plasma pressure data derived from ENA imaging, will also be discussed.