skip to main content


Title: Predicting Swarm Equatorial Plasma Bubbles via Machine Learning and Shapley Values
Abstract

In this study we present AI Prediction of Equatorial Plasma Bubbles (APE), a machine learning model that can accurately predict the Ionospheric Bubble Index (IBI) on the Swarm spacecraft. IBI is a correlation (R2) between perturbations in plasma density and the magnetic field, whose source can be Equatorial Plasma Bubbles (EPBs). EPBs have been studied for a number of years, but their day‐to‐day variability has made predicting them a considerable challenge. We build an ensemble machine learning model to predict IBI. We use data from 2014 to 2022 at a resolution of 1s, and transform it from a time‐series into a 6‐dimensional space with a corresponding EPBR2(0–1) acting as the label. APE performs well across all metrics, exhibiting a skill, association and root mean squared error score of 0.96, 0.98 and 0.08 respectively. The model performs best post‐sunset, in the American/Atlantic sector, around the equinoxes, and when solar activity is high. This is promising because EPBs are most likely to occur during these periods. Shapley values reveal that F10.7 is the most important feature in driving the predictions, whereas latitude is the least. The analysis also examines the relationship between the features, which reveals new insights into EPB climatology. Finally, the selection of the features means that APE could be expanded to forecasting EPBs following additional investigations into their onset.

 
more » « less
PAR ID:
10419871
Author(s) / Creator(s):
 ;  ;  ;  ;  ;  ;  ;  
Publisher / Repository:
DOI PREFIX: 10.1029
Date Published:
Journal Name:
Journal of Geophysical Research: Space Physics
Volume:
128
Issue:
6
ISSN:
2169-9380
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Jee, Geonhwa (Ed.)
    Electron density irregularities in the equatorial ionosphere at night are understood in terms of plasma bubbles, which are produced by the transport of low-density plasma from the bottomside of the F region to the topside. Equatorial plasma bubbles (EPBs) have been detected by various techniques on the ground and from space. One of the distinguishing characteristics of EPBs identified from long-term observations is the systematic seasonal and longitudinal variation of the EPB activity. Several hypotheses have been developed to explain the systematic EPB behavior, and now we have good knowledge about the key factors that determine the behavior. However, gaps in our understanding of the EPB climatology still remain primarily because we do not yet have the capability to observe seed perturbations and their growth simultaneously and globally. This paper reviews the occurrence climatology of EPBs identified from observations and the current understanding of its driving mechanisms. 
    more » « less
  2. Abstract

    The influence of atmospheric planetary waves on the occurrence of irregularities in the low latitude ionosphere is investigated using Whole Atmosphere Community Climate Model with thermosphere‐ionosphere eXtension (WACCM‐X) simulations and Global Observations of the Limb and Disk (GOLD) observations. GOLD observations of equatorial plasma bubbles (EPBs) exhibit a ∼6–8 day periodicity during January–February 2021. Analysis of WACCM‐X simulations, which are constrained to reproduce realistic weather variability in the lower atmosphere, reveals that this coincides with an amplification of the westward propagating wavenumber‐1 quasi‐six day wave (Q6DW) in the mesosphere and lower thermosphere (MLT). The WACCM‐X simulated Rayleigh‐Taylor (R‐T) instability growth rate, considered as a proxy of EPB occurrence, is found to exhibit a ∼6‐day periodicity that is coincident with the enhanced Q6DW in the MLT. Additional WACCM‐X simulations performed with fixed solar and geomagnetic activity demonstrate that the ∼6‐day periodicity in the R‐T instability growth rate is related to the forcing from the lower atmosphere. The simulations suggest that the Q6DW influences the day‐to‐day formation of EPBs through interaction with the migrating semidiurnal tide. This leads to periodic oscillations in the zonal winds, resulting in periodic variability in the strength of the prereversal enhancement, which influences the R‐T instability growth rate and EPBs. The results demonstrate that atmospheric planetary waves, and their interaction with atmospheric tides, can have a significant impact on the day‐to‐day variability of EPBs.

     
    more » « less
  3. Abstract

    The FORMOSAT‐3/COSMIC (F3/C) satellites are used to study the climatology of equatorial plasma bubbles (EPBs) during the low to moderate solar flux years (2008–2013). We use the F3/C total electron content to identify the presence of EPBs and investigate the background conditions for the initiation of EPBs. The results reveal that the EPB activities have strong solar dependence. The longitudinal and seasonal trends of EPBs are highly correlated to the angle between the dusk solar terminator and magnetic field lines near the magnetic equator. Asymmetries of EPBs between solstices and equinoxes exist and could be due partly to the asymmetry of equatorial ionization anomaly structures, which result in longitudinal differences as well. EPBs extend to higher altitudes and latitudes during the ascending phase of Solar Cycle 24 (2011–2013) due mainly to the increase of background electron density. However, an altitudinal asymmetry of EPBs occurs in moderate solar flux years, which is likely due to the suppression or lower growth and occurrence rates of EPBs. In addition to vertical drift, tidal forcing also contributes to the longitudinal and seasonal distributions of EPBs. Upwellings and precursor waves preceding the EPBs are observed climatologically, which likely play a vital role in initiating the EPBs. This study also reveals a vertical connection between the equatorial ionospheric irregularities and atmospheric forcing on a climatological basis.

     
    more » « less
  4. Abstract

    Predicting the daily variability of Equatorial Plasma Bubbles (EPBs) is an ongoing scientific challenge. Various methods for predicting EPBs have been developed, however, the research community is yet to scrutinize the methods for evaluating and comparing these prediction models/techniques. In this study, 12 months of co‐located GPS and UHF scintillation observations spanning South America, Atlantic/Western Africa, Southeast Asia, and Pacific sectors are used to evaluate the Generalized Rayleigh‐Taylor (R‐T) growth rates calculated from the Thermosphere Ionosphere Electrodynamics General Circulation Model (TIEGCM). Various assessment metrics are explored, including the use of significance testing on skill scores for threshold selection. The sensitivity of these skill scores to data set type (i.e., GPS versus UHF) and data set size (30, 50, 60, and 90 days/events) is also investigated. It is shown that between 50 and 90 days is required to achieve a statistically significant skill score. Methods for conducting model‐model comparisons are also explored, including the use of model “sufficiency.” However, it is shown that the results of model‐model comparisons must be carefully interpreted and can be heavily dependent on the data set used. It is also demonstrated that the observation data set must exhibit an appropriate level of daily EPB variability in order to assess the true strength of a given model/technique. Other limitations and considerations on assessment metrics and future challenges for EPB prediction studies are also discussed.

     
    more » « less
  5. Abstract

    This study develops a new Bubble Index to quantify the intensity of 2‐D postsunset equatorial plasma bubbles (EPBs) in the American/Atlantic sector, using Global‐scale Observations of the Limb and Disk (GOLD) nighttime data. A climatology and day‐to‐day variability analysis of EPBs is conducted based on the newly‐derived Bubble Index with the following results: (a) EPBs show considerable seasonal and solar activity dependence, with stronger (weaker) intensity around December (June) solstice and high (low) solar activity years. (b) EPBs exhibit opposite geomagnetic activity dependencies during different storm phases: EPBs are intensified concurrently with an increasing Kp, but are suppressed with high Kp occurring 3–6 hr earlier. (c) For the first time, we found that EPBs' day‐to‐day variation exhibited quasi‐3‐day and quasi‐6‐day periods. A coordinated analysis of Ionospheric Connection Explorer (ICON) winds and ionosonde data suggests that this multi‐day periodicity was related to the planetary wave modulation through the wind‐driven dynamo.

     
    more » « less