skip to main content

This content will become publicly available on March 24, 2023

Title: Identification and Classification of Relativistic Electron Precipitation at Earth Using Supervised Deep Learning
We show an application of supervised deep learning in space sciences. We focus on the relativistic electron precipitation into Earth’s atmosphere that occurs when magnetospheric processes (wave-particle interactions or current sheet scattering, CSS) violate the first adiabatic invariant of trapped radiation belt electrons leading to electron loss. Electron precipitation is a key mechanism of radiation belt loss and can lead to several space weather effects due to its interaction with the Earth’s atmosphere. However, the detailed properties and drivers of electron precipitation are currently not fully understood yet. Here, we aim to build a deep learning model that identifies relativistic precipitation events and their associated driver (waves or CSS). We use a list of precipitation events visually categorized into wave-driven events (REPs, showing spatially isolated precipitation) and CSS-driven events (CSSs, showing an energy-dependent precipitation pattern). We elaborate the ensemble of events to obtain a dataset of randomly stacked events made of a fixed window of data points that includes the precipitation interval. We assign a label to each data point: 0 is for no-events, 1 is for REPs and 2 is for CSSs. Only the data points during the precipitation are labeled as 1 or 2. By adopting a long more » short-term memory (LSTM) deep learning architecture, we developed a model that acceptably identifies the events and appropriately categorizes them into REPs or CSSs. The advantage of using deep learning for this task is meaningful given that classifying precipitation events by its drivers is rather time-expensive and typically must involve a human. After post-processing, this model is helpful to obtain statistically large datasets of REP and CSS events that will reveal the location and properties of the precipitation driven by these two processes at all L shells and MLT sectors as well as their relative role, thus is useful to improve radiation belt models. Additionally, the datasets of REPs and CSSs can provide a quantification of the energy input into the atmosphere due to relativistic electron precipitation, thus offering valuable information to space weather and atmospheric communities. « less
; ;
Award ID(s):
Publication Date:
Journal Name:
Frontiers in Astronomy and Space Sciences
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract

    Machine learning (ML) has been applied to space weather problems with increasing frequency in recent years, driven by an influx of in-situ measurements and a desire to improve modeling and forecasting capabilities throughout the field. Space weather originates from solar perturbations and is comprised of the resulting complex variations they cause within the numerous systems between the Sun and Earth. These systems are often tightly coupled and not well understood. This creates a need for skillful models with knowledge about the confidence of their predictions. One example of such a dynamical system highly impacted by space weather is the thermosphere, the neutral region of Earth’s upper atmosphere. Our inability to forecast it has severe repercussions in the context of satellite drag and computation of probability of collision between two space objects in low Earth orbit (LEO) for decision making in space operations. Even with (assumed) perfect forecast of model drivers, our incomplete knowledge of the system results in often inaccurate thermospheric neutral mass density predictions. Continuing efforts are being made to improve model accuracy, but density models rarely provide estimates of confidence in predictions. In this work, we propose two techniques to develop nonlinear ML regression models to predictmore »thermospheric density while providing robust and reliable uncertainty estimates: Monte Carlo (MC) dropout and direct prediction of the probability distribution, both using the negative logarithm of predictive density (NLPD) loss function. We show the performance capabilities for models trained on both local and global datasets. We show that the NLPD loss provides similar results for both techniques but the direct probability distribution prediction method has a much lower computational cost. For the global model regressed on the Space Environment Technologies High Accuracy Satellite Drag Model (HASDM) density database, we achieve errors of approximately 11% on independent test data with well-calibrated uncertainty estimates. Using an in-situ CHAllenging Minisatellite Payload (CHAMP) density dataset, models developed using both techniques provide test error on the order of 13%. The CHAMP models—on validation and test data—are within 2% of perfect calibration for the twenty prediction intervals tested. We show that this model can also be used to obtain global density predictions with uncertainties at a given epoch.

    « less
  2. In near-Earth space, the magnetosphere, energetic electrons (tens to thousands of kiloelectron volts) orbit around Earth, forming the radiation belts. When scattered by magnetospheric processes, these electrons precipitate to the upper atmosphere, where they deplete ozone, a radiatively active gas, modifying global atmospheric circulation. Relativistic electrons (those above a few hundred kiloelectron volts), can reach the lowest altitudes and have the strongest effects on the upper atmosphere; their loss from the magnetosphere is also important for space weather. Previous models have only considered magnetospheric scattering and precipitation of energetic electrons; atmospheric scattering of such electrons has not been adequately considered, principally due to lack of observations. Here we report the first observations of this process. We find that atmospherically-scattered energetic (relativistic) electrons form a low-intensity, persistent “drizzle”, whose integrated energy flux is comparable to (greater than) that of the more intense but ephemeral precipitation by magnetospheric scattering. Thus, atmospheric scattering of energetic electrons is important for global atmospheric circulation, radiation belt flux evolution, and the repopulation of the magnetosphere with lower-energy, secondary electrons.
  3. This paper presents observations of electromagnetic ion cyclotron (EMIC) waves from multiple data sources during the four Geospace Environment Modeling challenge events in 2013 selected by the Geospace Environment Modeling Quantitative Assessment of Radiation Belt Modeling focus group: 17 and 18 March (stormtime enhancement), 31 May to 2 June (stormtime dropout), 19 and 20 September (nonstorm enhancement), and 23–25 September (nonstorm dropout). Observations include EMIC wave data from the Van Allen Probes, Geostationary Operational Environmental Satellite, and Time History of Events and Macroscale Interactions during Substorms spacecraft in the near-equatorial magnetosphere and from several arrays of ground-based search coil magnetometers worldwide, as well as localized ring current proton precipitation data from low-altitude Polar Operational Environmental Satellite spacecraft. Each of these data sets provides only limited spatial coverage, but their combination shows consistent occurrence patterns and reveals some events that would not be identified as significant using near-equatorial spacecraft alone. Relativistic and ultrarelativistic electron flux observations, phase space density data, and pitch angle distributions based on data from the Relativistic Electron-Proton Telescope and Magnetic Electron Ion Spectrometer instruments on the Van Allen Probes during these events show two cases during which EMIC waves are likely to have played an important rolemore »in causing major flux dropouts of ultrarelativistic electrons, particularly near L* ~4.0. In three other cases, identifiable smaller and more short-lived dropouts appeared, and in five other cases, these waves evidently had little or no effect.« less
  4. Abstract

    Energetic electron precipitation from Earth’s outer radiation belt heats the upper atmosphere and alters its chemical properties. The precipitating flux intensity, typically modelled using inputs from high-altitude, equatorial spacecraft, dictates the radiation belt’s energy contribution to the atmosphere and the strength of space-atmosphere coupling. The classical quasi-linear theory of electron precipitation through moderately fast diffusive interactions with plasma waves predicts that precipitating electron fluxes cannot exceed fluxes of electrons trapped in the radiation belt, setting an apparent upper limit for electron precipitation. Here we show from low-altitude satellite observations, that ~100 keV electron precipitation rates often exceed this apparent upper limit. We demonstrate that such superfast precipitation is caused by nonlinear electron interactions with intense plasma waves, which have not been previously incorporated in radiation belt models. The high occurrence rate of superfast precipitation suggests that it is important for modelling both radiation belt fluxes and space-atmosphere coupling.

  5. Understanding the drivers of surface melting in West Antarctica is crucial for understanding future ice loss and global sea level rise. This study identifies atmospheric drivers of surface melt on West Antarctic ice shelves and ice sheet margins and relationships with tropical Pacific and high-latitude climate forcing using multidecadal reanalysis and satellite datasets. Physical drivers of ice melt are diagnosed by comparing satellite-observed melt patterns to anomalies of reanalysis near-surface air temperature, winds, and satellite-derived cloud cover, radiative fluxes, and sea ice concentration based on an Antarctic summer synoptic climatology spanning 1979–2017. Summer warming in West Antarctica is favored by Amundsen Sea (AS) blocking activity and a negative phase of the southern annular mode (SAM), which both correlate with El Niño conditions in the tropical Pacific Ocean. Extensive melt events on the Ross–Amundsen sector of the West Antarctic Ice Sheet (WAIS) are linked to persistent, intense AS blocking anticyclones, which force intrusions of marine air over the ice sheet. Surface melting is primarily driven by enhanced downwelling longwave radiation from clouds and a warm, moist atmosphere and by turbulent mixing of sensible heat to the surface by föhn winds. Since the late 1990s, concurrent with ocean-driven WAIS mass loss, summermore »surface melt occurrence has increased from the Amundsen Sea Embayment to the eastern Ross Ice Shelf. We link this change to increasing anticyclonic advection of marine air into West Antarctica, amplified by increasing air–sea fluxes associated with declining sea ice concentration in the coastal Ross–Amundsen Seas.

    « less