Abstract:Collisionless space plasma environments are typically characterized by distinct particle populations. Although moments of their velocity distribution functions help in distinguishing different plasma regimes, the distribution functions themselves provide more comprehensive information about the plasma state, especially at times when the distribution function includes non-thermal effects. Unlike moments, however, distribution functions are not easily characterized by a small number of parameters, making their classification more difficult to achieve. In order to perform this classification, we propose to distinguish between the different plasma regions by applying dimensionality reduction and clustering methods to electron distributions in pitch angle and energy space. We utilize four separate algorithms to achieve our plasma classifications: autoencoders, principal component analysis, mean shift, and agglomerative clustering. We test our classification algorithms by applying our scheme to data from the Cluster-Plasma Electron and Current Experiment instrument measured in the Earth’s magnetotail. Traditionally, it is thought that the Earth’s magnetotail is split into three different regions (the plasma sheet, the plasma sheet boundary layer, and the lobes), that are primarily defined by their plasma characteristics. Starting with the ECLAT database with associated classifications based on the plasma parameters, we identify eight distinct groups of distributions, that are dependent upon significantly more complex plasma and field dynamics. By comparing the average distributions as well as the plasma and magnetic field parameters for each region, we relate several of the groups to different plasma sheet populations, and the rest we attribute to the plasma sheet boundary layer and the lobes. We find clear distinctions between each of our classified regions and the ECLAT results. The automated classification of different regions in space plasma environments provides a useful tool to identify the physical processes governing particle populations in near-Earth space. These tools are model independent, providing reproducible results without requiring the placement of arbitrary thresholds, limits or expert judgment. Similar methods could be used onboard spacecraft to reduce the dimensionality of distributions in order to optimize data collection and downlink resources in future missions.

Clustering of Global Magnetospheric Observations

Unsupervised classification of simulated magnetospheric regions

Generalized Time‐Series Analysis for In Situ Spacecraft Observations: Anomaly Detection and Data Prioritization Using Principal Components Analysis and Unsupervised Clustering

Multispacecraft Measurements in the Magnetosphere

Use of twenty years CLUSTER/FGM data to observe the mean behavior of the magnetic field and current density of Earth's magnetosphere

Visualizing and Interpreting Unsupervised Solar Wind Classifications

Using Dimensionality Reduction and Clustering Techniques to Classify Space Plasma Regimes

MMS SITL Ground Loop: Automating the Burst Data Selection Process

A Database for Simultaneous Observations of the Earth's Magnetosheath by Cluster and MMS Between 2017 and 2021

The Cluster spacecrafts' view of the motion of the high-latitude magnetopause

Comparing 1-year GUMICS-4 simulations of the Terrestrial Magnetosphere with Cluster Measurements

Unsupervised classification of fully kinetic simulations of plasmoid instability using Self-Organizing Maps (SOMs)

Objectively Determining States of the Solar Wind Using Machine Learning

Recent Cluster Results on Cusp and Magnetopause Boundary Phenomena

The Cluster Magnetic Field Investigation: overview of in-flight performance and initial results

Use of Multi-Point Analysis and Modelling to Address Cross-Scale Coupling in Space Plasmas: Lessons from Cluster

Magnetotail Dipolarization and Associated Current Systems Observed by Cluster and Double Star

A Statistical Study of the Plasma Sheet in the Near and Middle Earth Magnetotail

Study of Waves in the Magnetotail Region with Cluster and DSP

Clustering of magnetic reconnection exhausts in the solar wind: An automated detection study

Classifying Magnetic Reconnection Regions Using k‐Means Clustering: Applications to Energy Partition