Data mining for extreme space weather
ROSES ID: NNH17ZDA001N Selection Year: 2017
Program Element: Focused Science Topic
Principal Investigator: Mikhail Sitnov
Affiliation(s): The Johns Hopkins Universtiy Applied Physics Laboratory
Stephens, Grant Killian Co-I Johns Hopkins University
Brandt, Pontus C. Co-I Johns Hopkins University
Merkin, Viacheslav G. Co-I Johns Hopkins University
Science goals and objectives: The overarching goal is to reveal distinctive features of the Earth's magnetosphere during extreme events (EEs) through empirical reconstruction of the geomagnetic field, electric currents and plasma pressure for superstorms (Dst index<-300 nT), taking into account their statistical peculiarity as EEs. We will compare distributions of the magnetospheric parameters with similar distributions for weaker storms and first-principle simulations to grasp the distinctions of EEs and their underlying mechanisms. Since EEs correspond to tails of the corresponding data distributions, their empirical pictures are strongly biased toward weaker events and their errors are large, making their interpretation particularly challenging. At the same time, the error analysis can be used to improve the empirical picture of EEs, and this improvement is also one of the main objectives of this study, which is guided therefore by following science questions:
1) What are the distinctive features of spatial distribution and temporal evolution of the magnetic field, electric currents and plasma pressure during superstorms?
2) What are the key biases and uncertainties of empirical reconstructions of these quantities for EE activity level, and how can they be used to improve the empirical picture?
Methodology: The reconstruction will be made using the nearest neighbor (NN) data mining algorithm. In this approach, the state of the magnetosphere as well as its evolution are parametrized by the global activity parameters, Dst or Sym-H index, the solar wind electric field and their time derivatives. Then the magnetic field for a query event can be reconstructed using only a small subset of the whole historical database of magnetometer records. The resulting empirical picture, and in particular, distributions of the force-balanced plasma pressure, will be compared with pressure distributions retrieved from Energetic Neutral Atom (ENA) observations for EEs observed by the IMAGE mission in the period 2000-2004, to adjust the location of the pressure peak obtained by the NN method. Eventually, the amplitudes of empirical distributions will be adjusted based on the error and bias analyses to match the observed Sym-H values for the smallest possible NN bins. The empirical analysis of EEs will be complemented by similar studies of strong storms with Dst < -200 nT, including the strongest storms in the Van Allen Probes mission era, to validate pressure distributions using direct particle measurements and to improve the statistics of the pressure peak locations. The empirical model will be further extended to older superstorms in the period 1957-1991, when no solar wind monitors were operational, and hence no event-specific first-principles modeling is possible.
To understand the key physical processes specific for superstorms, their empirical picture will be compared with global simulations of the magnetosphere using the Lyon-Fedder-Mobarry (LFM) MHD model coupled with the kinetic Rice Convection model (RCM) of the ring current evolution. Further the RCM pressure will be replaced by its empirical analog inferred from the magnetic field distributions using the quasi-static force balance equation.
Proposed Contributions to the Focus Team Effort: The project contributes to FST #4 by providing an observational basis for identification of the key physical processes during EEs. The potential contributions to the FST's team effort will be empirical, first-principles and combined pictures of the magnetic field, currents and plasma pressure. These results will mark the milestones of the project. They can be used to improve understanding of the EE physics and provide direct assessment of the key Space Weather factors, such as the geomagnetically induced currents. The metrics of success will be derived from comparisons with ENA and in situ measurements.
Export to PDF