New model fit functions of the plasmapause location determined using THEMIS observations during the ascending phase of Solar Cycle 24

Abstract

It is well known that the plasmapause is influenced by the solar wind and magnetospheric conditions. Empirical models of its location have been previously developed such as those by O'Brien and Moldwin (2003) and Larsen et al. (2006). In this study, we identified the locations of the plasmapause using the plasma density data obtained from the Time History of Events and Macroscale Interactions during Substorms (THEMIS) satellites. We used the data for the period (2008–2012) corresponding to the ascending phase of Solar Cycle 24. Our database includes data from over a year of unusually weak solar wind conditions, correspondingly covering the plasmapause locations in a wider L range than those in previous studies. It also contains many coronal hole stream intervals during which the plasmasphere is eroded and recovers over a timescale of several days. The plasmapause was rigorously determined by requiring a density gradient by a factor of 15 within a radial distance of 0.5 L. We first determined the statistical correlation of the plasmapause locations with several solar wind parameters as well as geomagnetic indices. We found that the plasmapause locations are well correlated with the solar wind speed and the interplanetary magnetic field Bz, therefore the y component of the convective electric field, and some energy coupling functions such as the well‐known Akasofu's epsilon parameter. The plasmapause locations are also highly correlated with the geomagnetic indices, Dst, AE, and Kp, as recognized previously. Finally, we suggest new model fit functions for the plasmapause locations in terms of the solar wind parameters and geomagnetic indices. When applied to a new data interval outside the model training interval, our model fit functions work better than existing ones. The new model fit functions developed here extend the range of conditions from those used in previous works.

A new solar wind‐driven global dynamic plasmapause model: 1. Database and statistics

“…[]. The calculated electron densities are associated with an error of a factor of 2, which is smaller than the density drop around the plasmapause, and have been widely used to identify the plasmapause locations [ Li et al ., ; Cho et al ., ; Liu et al ., ]. An example of the electron density profile and corresponding satellite potential is shown in Figures a and b.…”

Development of a 3‐D Plasmapause Model With a Back‐Propagation Neural Network

“…Only in situ measurements are used (Table 1). A plasmapause crossing is defined as a plasma density decrease by a factor of 5 or more within 0.5 Re (Carpenter & Anderson, 1992;Cho et al, 2015;Liu et al, 2015;Moldwin et al, 2002) and is located at the middle of the density drop. The first type is plasma wave instruments (e.g., Akebono, Cluster, IMAGE, Polar, and VAP).…”

“…Generally, all these models were constructed with single-or multiple-parameter fitting. Since the direct driver of the variations in geospace is the external solar wind, the solar wind and interplanetary magnetic field (IMF) parameters were then introduced into the models (e.g., Bandić et al, 2016;Cho et al, 2015;Larsen et al, 2007;Verbanac et al, 2015). Following this scheme, several works have used other geomagnetic indices such as Dst, SYM-H, and AE, to construct models (Moldwin et al, 2002;O'Brien & Moldwin, 2003).…”

Narrow-band emission with 0.5 to 3.5 Hz varying frequency in the background of the main phase of the 17 March 2013 magnetic storm

“…At the same time, the plasmasphere started to be affected by accelerated moderate-energy ions from the geomagnetic tail. Further, taking into account the conclusions drawn in [Goldstein et al, 2005;Dandouras et al, 2009;Liu et al, 2015;Cho et al, 2015], it is fair to assume that in subsequent hours, on the one hand, the plasmapause approached Earth and by 21 UT reached L p =2.5-3 in its near-midnight part; on the other hand, the successive injections produced particle fluxes of increasing energy, which penetrated deeper into the magnetosphere, reaching outer areas of the plasmasphere.…”

“…Numerous empirical models predicting the position of the plasmapause [Carpenter, Anderson, 1992;Moldwin et al, 2002;Liu et al, 2015;Cho et al, 2015;Verbanac et al, 2015] work badly under high magnetic activity. It suffices to say that substituting real solar wind parameters and geomagnetic indices for March 17, 2013 in the model [Verbanac et al, 2015] gives negative L p values for night hours.…”