NSF Award
0418857
This project will further develop a model originally funded with earlier NSF support to a space weather model for the fluctuations of ground magnetic field. The current model is a linear dynamical model trained on historic measurements taken by the Finnish magnetometer array, IMAGE, the geomagnetic PC index, and the recent history of the solar wind inputs. These elements represent the state of geomagnetic activity and of the global magnetosphere, and a measure of the energy available through dayside magnetic reconnection. The model produces an estimate of the horizontal (2D) field at latitudes 55-90 degrees. The dynamics of the field are approximated as oscillation and growth/decay and they are driven by the PC index and solar wind inputs. The proposed model will be trained to predict the full 3D ground magnetic field. A similar model will be separately developed to predict the field fluctuations in terms of the time derivatives of the magnetic field. Time derivatives are of particular importance because geomagnetically induced currents, which can effect the electric power distribution grid, are driven by sudden changes in the magnetic field. The models will be nonlinear, meaning that their coefficients will be parameterized by the geomagnetic activity level. The transition from linear to nonlinear dynamics will increase the prediction accuracy. The effects of the solar wind and interplanetary magnetic field (IMF) inputs on prediction accuracy will be examined. The solar wind-magnetosphere coupling process depends on local time as well as on latitude. Each location of the model can be driven by a distinct solar wind/IMF input function. The proposed model can dynamically adapt to changing solar wind conditions, but it is important that it be regularly corrected by the correct present state of geomagnetic conditions. Data assimilation will be used to ingest real-time geomagnetic measurements. Because the model can be written as a filter for the local field (or its fluctuations), it will be extended to include a Kalman filter. The Kalman component will adjust the state of the model so that it rapidly converges on realistic conditions at the points of measurement. A dynamic model for the ground magnetic field has several important applications. First, understanding the geomagnetic response and its dependence on solar wind input and the magnetospheric configuration is an important question in magnetospheric plasma physics. The model can be used as a tool to quantify this complex coupling as a function of interplanetary conditions and the magnetospheric state. In addition, some aspects of the model can be used as part of student training on magnetospheric and ionospheric plasma physics.
