4:15 PM - 4:30 PM
[MGI29-10] Preliminary results of the data assimilation applied to the global MHD simulation code toward reanalysis of the space weather phenomena
Keywords:Space Weather, Data Assimilation, global MHD simulation
It is almost impossible to obtain an accurate 3-D image of the magnetospheric phenomena only from the observations because the direct (in-situ) observations in the magnetosphere are quite sparse. Numerical simulations that accurately solve the physical first principles are powerful tools for studying phenomena occurring in the magnetosphere. When we can provide the grid-point-values of the plasmas in the magnetosphere-ionosphere system obtained from the simulation to the public, it may be very useful to the scientific society. So, the magnetosphere-ionosphere global simulation can be used as a useful tool for magnetospheric physics. The final goal of the present project is to produce the grid-point values of the magnetosphere-ionosphere plasma processes and to make them publicly available.
For the purposes stated above, we, at first, improved the conventional global magnetosphere-ionosphere simulation code (this is the REPPU code by Tanaka (2015)) to include both the effect of the inclined rotation axis of the Earth and the effect of the discrepancy between the rotational axis and the magnetic axis. In addition, the improved REPPU code utilizes all three components of the IMF and solar wind velocity. Next, we evaluated how the improved REPPU code reproduces the observations by using SuperDARN electric potential data, AMPERE field-aligned current data, and AE indices. Thus, we obtain the following results;The correlation coefficient between the calculated electric potential in the polar region and the superDARN potential is about 0.8.The simulation almost reproduces the field-aligned current pattern observed by the AMPERE. However, the field-aligned current intensity and location are not wholly the same between the simulation and the observation. Therefore, the correlation between the simulation and the observation becomes low.Temporal variations of the simulated AE indices are parallel to the observed one.As a result, the improved REPPU code reasonably reproduces plasma processes of the real magnetosphere-ionosphere phenomena. To improve the accuracy, we need to employ the data assimilation technique.
The improved REPPU code assumes the ionospheric conductivities controlled by electron precipitation, the ion precipitation, and the ratio between the Pedersen conductivity and the Hall conductivity. The ionospheric electric potential and the current in the ionosphere in the simulation are dependent on arbitrary factors that determine the ionospheric conductivities. Here, we start to determine the factors by the data assimilation. First, as we notice that the AE indices change sensitively to changes in the factors, the ensemble variational method is used to determine the factors from only the AE indices. However, assimilation does not give reasonable answers because the factors determined are far from the expected values. So, we start over to determine the factors from the assimilation technique by using not only the AE indices but also the SupeDARN electric potential. This time, we obtained the factors that seem to be reasonable. The details are reported in the talk.
In the talk, we will present an example of event analysis of the magnetic storm in 20150623-24 to demonstrate that the “reanalysis data” by the improved REPPU code are useful to space weather research. Our future goal is to provide a database of the reanalysis data for the space weather.
References
Tanaka, T. (2015), Substorm auroral dynamics reproduced by the advanced global M-I coupling simulation, In Auroral dynamics and space weather, Geophys. Monogr. Ser., vol. 215, edited by Y. Zhang and L. J. Paxton, p. 177, doi: 10.1002/9781118978719, AGU, Washington D. C.
For the purposes stated above, we, at first, improved the conventional global magnetosphere-ionosphere simulation code (this is the REPPU code by Tanaka (2015)) to include both the effect of the inclined rotation axis of the Earth and the effect of the discrepancy between the rotational axis and the magnetic axis. In addition, the improved REPPU code utilizes all three components of the IMF and solar wind velocity. Next, we evaluated how the improved REPPU code reproduces the observations by using SuperDARN electric potential data, AMPERE field-aligned current data, and AE indices. Thus, we obtain the following results;The correlation coefficient between the calculated electric potential in the polar region and the superDARN potential is about 0.8.The simulation almost reproduces the field-aligned current pattern observed by the AMPERE. However, the field-aligned current intensity and location are not wholly the same between the simulation and the observation. Therefore, the correlation between the simulation and the observation becomes low.Temporal variations of the simulated AE indices are parallel to the observed one.As a result, the improved REPPU code reasonably reproduces plasma processes of the real magnetosphere-ionosphere phenomena. To improve the accuracy, we need to employ the data assimilation technique.
The improved REPPU code assumes the ionospheric conductivities controlled by electron precipitation, the ion precipitation, and the ratio between the Pedersen conductivity and the Hall conductivity. The ionospheric electric potential and the current in the ionosphere in the simulation are dependent on arbitrary factors that determine the ionospheric conductivities. Here, we start to determine the factors by the data assimilation. First, as we notice that the AE indices change sensitively to changes in the factors, the ensemble variational method is used to determine the factors from only the AE indices. However, assimilation does not give reasonable answers because the factors determined are far from the expected values. So, we start over to determine the factors from the assimilation technique by using not only the AE indices but also the SupeDARN electric potential. This time, we obtained the factors that seem to be reasonable. The details are reported in the talk.
In the talk, we will present an example of event analysis of the magnetic storm in 20150623-24 to demonstrate that the “reanalysis data” by the improved REPPU code are useful to space weather research. Our future goal is to provide a database of the reanalysis data for the space weather.
References
Tanaka, T. (2015), Substorm auroral dynamics reproduced by the advanced global M-I coupling simulation, In Auroral dynamics and space weather, Geophys. Monogr. Ser., vol. 215, edited by Y. Zhang and L. J. Paxton, p. 177, doi: 10.1002/9781118978719, AGU, Washington D. C.