Title: Automating Burst Data Management to Simplify Mission Operations

Abstract:

Global-scale energy flow throughout Earth’s magnetosphere is catalyzed by processes that occur at Earth’s magnetopause (MP) in the electron diffusion region (EDR) of magnetic reconnection. Until the launch of the Magnetospheric Multiscale (MMS) mission only rare, fortuitous circumstances permitted a glimpse of the electron dynamics that break magnetic field lines and energize plasma. MMS employs automated burst triggers onboard the spacecraft and a Scientist-in-the-Loop (SITL) on the ground to select intervals likely to contain diffusion regions. Only low-resolution survey data is available to the SITL, which is insufficient to resolve electron dynamics. A strategy for the SITL, then, is to select all MP crossings. This has resulted in >35 potential MP EDR encounters but is labor- and resource-intensive; after manual classification, just ~0.7% of MP crossings, or ~0.0001% of the mission lifetime, during MMS’s first two years contained an EDR. We introduce two machine learning models to detect MP crossings and automate the SITL classification process: 1) a hierarchical Bayesian mixture model (HBM) with linear and auto regressive components, and 2) a limited short term memory (LSTM) recurrent neural network. The HBM selects twice as many magnetopause regions as a boosted regression tree model without significant over selection, achieving a 31% true positive rate and 93% true negative rate. The LSTM has been implemented in the MMS data stream to provide automated predictions to the SITL. Both methods are designed to make mission operations more efficient by consolidating manual classification processes into automated routines.