Probabilistic Forecasting of Ground Magnetic Perturbation Spikes at Mid‐Latitude Stations

Abstract The prediction of large fluctuations in the ground magnetic field (dB/dt) is essential for preventing damage from Geomagnetically Induced Currents. Directly forecasting these fluctuations has proven difficult, but accurately determining the risk of extreme events can allow for the worst of...

Full description

Saved in:
Bibliographic Details
Main Authors: Michael Coughlan, Amy Keesee, Victor Pinto, Raman Mukundan, José Paulo Marchezi, Jeremiah Johnson, Hyunju Connor, Don Hampton
Format: Article
Language:English
Published: Wiley 2023-06-01
Series:Space Weather
Subjects:
Online Access:https://doi.org/10.1029/2023SW003446
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850094563117498368
author Michael Coughlan
Amy Keesee
Victor Pinto
Raman Mukundan
José Paulo Marchezi
Jeremiah Johnson
Hyunju Connor
Don Hampton
author_facet Michael Coughlan
Amy Keesee
Victor Pinto
Raman Mukundan
José Paulo Marchezi
Jeremiah Johnson
Hyunju Connor
Don Hampton
author_sort Michael Coughlan
collection DOAJ
description Abstract The prediction of large fluctuations in the ground magnetic field (dB/dt) is essential for preventing damage from Geomagnetically Induced Currents. Directly forecasting these fluctuations has proven difficult, but accurately determining the risk of extreme events can allow for the worst of the damage to be prevented. Here we trained Convolutional Neural Network models for eight mid‐latitude magnetometers to predict the probability that dB/dt will exceed the 99th percentile threshold 30–60 min in the future. Two model frameworks were compared, a model trained using solar wind data from the Advanced Composition Explorer (ACE) satellite, and another model trained on both ACE and SuperMAG ground magnetometer data. The models were compared to examine if the addition of current ground magnetometer data significantly improved the forecasts of dB/dt in the future prediction window. A bootstrapping method was employed using a random split of the training and validation data to provide a measure of uncertainty in model predictions. The models were evaluated on the ground truth data during eight geomagnetic storms and a suite of evaluation metrics are presented. The models were also compared to a persistence model to ensure that the model using both datasets did not over‐rely on dB/dt values in making its predictions. Overall, we find that the models using both the solar wind and ground magnetometer data had better metric scores than the solar wind only and persistence models, and was able to capture more spatially localized variations in the dB/dt threshold crossings.
format Article
id doaj-art-6f0873ddbf23429bb17e1134ea640db8
institution DOAJ
issn 1542-7390
language English
publishDate 2023-06-01
publisher Wiley
record_format Article
series Space Weather
spelling doaj-art-6f0873ddbf23429bb17e1134ea640db82025-08-20T02:41:39ZengWileySpace Weather1542-73902023-06-01216n/an/a10.1029/2023SW003446Probabilistic Forecasting of Ground Magnetic Perturbation Spikes at Mid‐Latitude StationsMichael Coughlan0Amy Keesee1Victor Pinto2Raman Mukundan3José Paulo Marchezi4Jeremiah Johnson5Hyunju Connor6Don Hampton7Department of Physics & Astronomy University of New Hampshire Durham NH USADepartment of Physics & Astronomy University of New Hampshire Durham NH USADepartamento de Fisica Universidad de Santiago de Chile Santiago ChileDepartment of Physics & Astronomy University of New Hampshire Durham NH USADepartment of Physics & Astronomy University of New Hampshire Durham NH USADepartment of Electrical & Computer Engineering University of New Hampshire Manchester NH USANASA Goddard Space Flight Center Greenbelt MD USAGeophysical Institute University of Alaska Fairbanks Fairbanks AK USAAbstract The prediction of large fluctuations in the ground magnetic field (dB/dt) is essential for preventing damage from Geomagnetically Induced Currents. Directly forecasting these fluctuations has proven difficult, but accurately determining the risk of extreme events can allow for the worst of the damage to be prevented. Here we trained Convolutional Neural Network models for eight mid‐latitude magnetometers to predict the probability that dB/dt will exceed the 99th percentile threshold 30–60 min in the future. Two model frameworks were compared, a model trained using solar wind data from the Advanced Composition Explorer (ACE) satellite, and another model trained on both ACE and SuperMAG ground magnetometer data. The models were compared to examine if the addition of current ground magnetometer data significantly improved the forecasts of dB/dt in the future prediction window. A bootstrapping method was employed using a random split of the training and validation data to provide a measure of uncertainty in model predictions. The models were evaluated on the ground truth data during eight geomagnetic storms and a suite of evaluation metrics are presented. The models were also compared to a persistence model to ensure that the model using both datasets did not over‐rely on dB/dt values in making its predictions. Overall, we find that the models using both the solar wind and ground magnetometer data had better metric scores than the solar wind only and persistence models, and was able to capture more spatially localized variations in the dB/dt threshold crossings.https://doi.org/10.1029/2023SW003446geomagnetically induced currentsground magnetic disturbancemachine learningconvolutional neural networksexplainable machine learningspace weather
spellingShingle Michael Coughlan
Amy Keesee
Victor Pinto
Raman Mukundan
José Paulo Marchezi
Jeremiah Johnson
Hyunju Connor
Don Hampton
Probabilistic Forecasting of Ground Magnetic Perturbation Spikes at Mid‐Latitude Stations
Space Weather
geomagnetically induced currents
ground magnetic disturbance
machine learning
convolutional neural networks
explainable machine learning
space weather
title Probabilistic Forecasting of Ground Magnetic Perturbation Spikes at Mid‐Latitude Stations
title_full Probabilistic Forecasting of Ground Magnetic Perturbation Spikes at Mid‐Latitude Stations
title_fullStr Probabilistic Forecasting of Ground Magnetic Perturbation Spikes at Mid‐Latitude Stations
title_full_unstemmed Probabilistic Forecasting of Ground Magnetic Perturbation Spikes at Mid‐Latitude Stations
title_short Probabilistic Forecasting of Ground Magnetic Perturbation Spikes at Mid‐Latitude Stations
title_sort probabilistic forecasting of ground magnetic perturbation spikes at mid latitude stations
topic geomagnetically induced currents
ground magnetic disturbance
machine learning
convolutional neural networks
explainable machine learning
space weather
url https://doi.org/10.1029/2023SW003446
work_keys_str_mv AT michaelcoughlan probabilisticforecastingofgroundmagneticperturbationspikesatmidlatitudestations
AT amykeesee probabilisticforecastingofgroundmagneticperturbationspikesatmidlatitudestations
AT victorpinto probabilisticforecastingofgroundmagneticperturbationspikesatmidlatitudestations
AT ramanmukundan probabilisticforecastingofgroundmagneticperturbationspikesatmidlatitudestations
AT josepaulomarchezi probabilisticforecastingofgroundmagneticperturbationspikesatmidlatitudestations
AT jeremiahjohnson probabilisticforecastingofgroundmagneticperturbationspikesatmidlatitudestations
AT hyunjuconnor probabilisticforecastingofgroundmagneticperturbationspikesatmidlatitudestations
AT donhampton probabilisticforecastingofgroundmagneticperturbationspikesatmidlatitudestations