The Vanishing Martian Magnetic Pile-up Region: Probing Radial IMF Causality using MAVEN Measurements

The complex and variable interaction between the conducting Martian ionosphere and the incoming solar wind causes the draping of interplanetary magnetic field (IMF) lines around Mars, giving rise to a weak induced magnetosphere (IM) despite the planet’s lack of a global intrinsic magnetic field. The weak magnetic field is a result of induced ionospheric currents and a nearly perpetual dynamic feature of the Martian magnetic topology. This draping of IMF lines can be observed at the crossing of the magnetic pile-up plasma boundary (MPB), defined by a characteristic increase in magnetic field magnitude and attenuation in fluctuations, along with a significant decrease in the density of 1 keV protons. Alternatively known as the induced magnetosphere boundary (IMB), the MPB marks the separation between the magnetosheath and the Martian induced magnetosphere. We visually inspected the magnetic field and plasma data collected by the MAG, SWEA, and SWIA instruments on the Mars Atmosphere and Volatile EvolutioN (MAVEN) spacecraft in 2018, and found unusual isolated occurrences of unchanged or dropping magnetic field intensity during high solar zenith angle MPB crossings. These observations are interpreted as “disappearing” MPR phenomena as the magnetic pile-up signature is not observed, suggesting a reduced pile-up of IMF around Mars. Previous observations on Venus have attributed absences in the dayside IM to radial IMF orientation during extreme solar wind conditions, hampering magnetic draping as the flow of solar wind is close to aligned with the IMF (Zhang et al., 2009). Preliminary analysis of hourly cadence solar wind predictions reveals that this may also be true at Mars, potentially explaining some of the events. However, the Martian magnetosphere is shown to respond to solar wind fluctuations in a matter of minutes, making it important to explore higher resolution data and examine fluctuations on that scale to establish correlations and determine if this is an externally driven phenomenon or driven within the system itself instead. We will also discuss the concurrent development of a recurrent neural network (RNN) with long short-term memory (LSTM) architecture, which will aid in expanding the non-pile-up dataset to the 10 years of MAVEN data for a more robust investigation into the origin of MPR “disappearance”.