Premium
A 2‐D empirical plasma sheet pressure model for substorm growth phase using the Support Vector Regression Machine
Author(s) -
Yue Chao,
Wang ChihPing,
Lyons Larry,
Wang Yongli,
Hsu TungShin,
Henderson Michael,
Angelopoulos Vassilis,
Lui A. T. Y.,
Nagai Tsugunobu
Publication year - 2015
Publication title -
journal of geophysical research: space physics
Language(s) - English
Resource type - Journals
eISSN - 2169-9402
pISSN - 2169-9380
DOI - 10.1002/2014ja020787
Subject(s) - substorm , plasma sheet , solar wind , sunspot , phase (matter) , dynamic pressure , plasma , kinetic energy , physics , mechanics , astrophysics , magnetosphere , classical mechanics , nuclear physics , quantum mechanics , magnetic field
The plasma sheet pressure and its spatial structure during the substorm growth phase are crucial to understanding the development and initiation of substorms. In this paper, we first statistically analyzed the growth phase pressures using Geotail and Time History of Events and Macroscale Interactions during Substorms data and identified that solar wind dynamic pressure ( P SW ), energy loading, and sunspot number as the three primary factors controlling the growth phase pressure change. We then constructed a 2‐D equatorial empirical pressure model and an error model within r ≤ 20 R E using the Support Vector Regression Machine with the three factors as input. The model predicts the plasma sheet pressure accurately with median errors of 5%, and predicted pressure gradients agree reasonably well with observed gradients obtained from two‐probe measurements. The model shows that pressure increases linearly as P SW increases, and the P SW effect is stronger under lower energy loading. However, the pressure responses to energy loading and sunspot number are nonlinear. The pressure increases first with increasing loading or sunspot number, then remains relatively constant after reaching a peak value at ~8000 kV min loading or sunspot number of ~80. The loading effect is stronger when P SW is lower and the pressure variations are stronger near midnight than away from midnight. The sunspot number effect is clearer at smaller r . The pressure model can also be applied to understand the pressure changes observed during a substorm event by providing evaluations of the effects of energy loading and P SW , as well as the temporal and spatial effects along the spacecraft trajectory.