skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: Topside Electron Density Modeling Using Neural Network and Empirical Model Predictions
Abstract We model the electron density in the topside of the ionosphere with an improved machine learning (ML) model and compare it to existing empirical models, specifically the International Reference Ionosphere (IRI) and the Empirical‐Canadian High Arctic Ionospheric Model (E‐CHAIM). In prior work, an artificial neural network (NN) was developed and trained on two solar cycles worth of Defense Meteorological Satellite Program data (113 satellite‐years), along with global drivers and indices to predict topside electron density. In this paper, we highlight improvements made to this NN, and present a detailed comparison of the new model to E‐CHAIM and IRI as a function of location, geomagnetic condition, time of year, and solar local time. We discuss precision and accuracy metrics to better understand model strengths and weaknesses. The updated neural network shows improved mid‐latitude performance with absolute errors lower than the IRI by 2.5 × 109to 2.5 × 1010e−/m3, modestly improved performance in disturbed geomagnetic conditions with absolute errors reduced by about 2.5 × 109 e−/m3at high Kp compared to the IRI, and high Kp percentage errors reduced by >50% when compared to E‐CHAIM.  more » « less
Award ID(s):
2139916 2221765
PAR ID:
10488843
Author(s) / Creator(s):
;
Publisher / Repository:
Wiley
Date Published:
Journal Name:
Space Weather
Volume:
21
Issue:
12
ISSN:
1542-7390
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; ; ; ; ; ; ; ; et al (, Space Weather)
    Abstract The upper boundary height of the traditional community general circulation model of the ionosphere‐thermosphere system is too low to be applied to the topside ionosphere/thermosphere study. In this study, the National Center for Atmospheric Research Thermosphere‐Ionosphere‐Electrodynamics General Circulation Model (NCAR‐TIEGCM) was successfully extended upward by four scale heights from 400–600 km to 700–1,200 km depending on solar activity, named TIEGCM‐X. The topside ionosphere and thermosphere simulated by TIEGCM‐X agree well with the observations derived from a topside sounder and satellite drag data. In addition, the neutral density, temperature, and electron density simulated by TIEGCM‐X are morphologically consistent with the NCAR‐TIEGCM simulations before extension. The latitude‐altitude distribution of the equatorial ionization anomaly derived from TIEGCM‐X is more reasonable. During geomagnetic storm events, the thermospheric responses of TIEGCM‐X are similar to NCAR‐TIEGCM. However, the ionospheric storm effects in TIEGCM‐X are stronger than those in NCAR‐TIEGCM and are even opposites at some middle and low latitudes due to the presence of more closed magnetic field lines. Defense Meteorological Satellite Program observations prove that the ionospheric storm effect of TIEGCM‐X is more reasonable. The well‐validated TIEGCM‐X has significant potential applications in ionospheric/thermospheric studies, such as the responses to storms, low‐latitude dynamics, and data assimilation. 
    more » « less
  2. ; ; ; ; (, Journal of Geophysical Research: Space Physics)
    Abstract E‐region models have traditionally underestimated the ionospheric electron density. We believe that this deficiency can be remedied by using high‐resolution photoabsorption and photoionization cross sections in the models. Deep dips in the cross sections allow solar radiation to penetrate deeper into the E‐region producing additional ionization. To validate our concept, we perform a study of model electron density profiles (EDPs) calculated using the Atmospheric Ultraviolet Radiance Integrated Code (AURIC; D. Strickland et al., 1999,https://doi.org/10.1016/s0022-4073(98)00098-3) in the E‐region of the terrestrial ionosphere. We compare AURIC model outputs using new high‐resolution photoionization and photoabsorption cross sections, and solar spectral irradiances during low solar activity with incoherent scatter radar (ISR) measurements from the Arecibo and Millstone Hills observatories, Constellation Observing System for Meteorology Ionosphere and Climate (COSMIC‐1) observations, and outputs from empirical models (IRI‐2016 and FIRI‐2018). AURIC results utilizing the new high‐resolution cross sections reveal a significant difference to model outputs calculated with the low‐resolution cross sections currently used. Analysis of AURIC EDPs using the new high‐resolution data indicate fair agreement with ISR measurements obtained at various times at Arecibo but very good agreement with Millstone Hills ISR observations from ∼96–140 km. However, discrepancies in the altitude of the E‐region peak persist. High‐resolution AURIC calculations are in agreement with COSMIC‐1 observations and IRI‐2016 model outputs between ∼105 and 140 km while FIRI‐2018 outputs underestimate the EDP in this region. Overall, AURIC modeling shows increased E‐region electron densities when utilizing high‐resolution cross sections and high‐resolution solar irradiances, and are likely to be the key to resolving the long standing data‐model discrepancies. 
    more » « less
  3. ; ; ; ; (, Journal of Geophysical Research: Space Physics)
    Abstract Local empirical models of the F2 layer peak electron density (NmF2) are developed for 43 low‐ middle latitude ionosonde stations using auto‐scaled data from Lowell GIRO data center and manually scaled data from World Data Center for Ionosphere and Space Weather. Data coverage at these stations ranges from a few years to up to 6 decades. Flare Irradiance Spectral Model index version 2 (FISM2) and ap3 index are used to parametrize the solar extreme ultraviolet (EUV) flux and geomagnetic activity dependence of NmF2. Learning curves suggest that approximately 8 years of data coverage is required to constrain the solar activity dependence of NmF2. Output of local models altogether captures well known anomalies of the F2 ionospheric layer. Performance metrics demonstrate that the model parametrized using FISM2 has better accuracy than a similarly parametrized model with F10.7, as well as than the IRI‐2020 model. Skill score metrics indicate that the FISM2 based model outperforms F10.7 model at all solar activity levels. The improved accuracy of model with FISM2 over F10.7 is due to better representation of solar rotation by FISM2, and due to its performance at solar extremum. Application of singular spectrum analysis to model output reveals that solar rotation contributes to about 2%–3% of the variance in NmF2 data and FISM2 based model, while F10.7 based models overestimate the strength of solar rotation to be at 4%–7%. At solar extremum, both F10.7‐based model and IRI‐2020 tend to overestimate the NmF2 while FISM2 provides the most accurate prediction out of three. 
    more » « less
  4. ; ; ; ; ; ; ; ; ; ; et al (, Radio Science)
    Abstract Slant absolute total electron content (TEC) is observed by the Formosa Satellite‐7/Constellation Observing System for Meteorology, Ionosphere, and Climate‐2 (FORMOSAT‐7/COSMIC‐2, F7/C2) Tri‐GNSS Radio Occultation System (TGRS) instrument. We present details of the data processing algorithms, validation, and error assessment for the F7/C2 global positioning system (GPS) absolute TEC observations. The data processing includes estimation and application of solar panel dependent pseudorange multipath maps, phase to pseudorange leveling, and estimation of separate L1C‐L2C and L1C‐L2P receiver differential code biases. We additionally perform a validation of the F7/C2 GPS absolute TEC observations through comparison with colocated, independent, TEC observations from the Swarm‐B satellite. Based on this comparison, we conclude that the accuracy of the F7/C2 GPS absolute TEC observations is less than 3.0 TEC units. Results are also presented that illustrate the suitability of the F7/C2 GPS absolute TEC observations for studying the climatology and variability of the topside ionosphere and plasmasphere (i.e., altitudes above the F7/C2 orbit of550 km). These results demonstrate that F7/C2 provides high quality GPS absolute TEC observations that can be used for ionosphere‐thermosphere data assimilation as well as scientific studies of the topside ionosphere and plasmasphere. 
    more » « less
  5. ; ; ; ; (, Space Weather)
    Abstract Specification and forecast of ionospheric parameters, such as ionospheric electron density (Ne), have been an important topic in space weather and ionospheric research. Neural networks (NNs) emerge as a powerful modeling tool forNeprediction. However, heavy manual adjustments are time consuming to determine the optimal NN structures. In this work, we propose to use neural architecture search (NAS), an automatic machine learning method, to mitigate this problem. NAS aims to find the optimal network structure through the alternate optimization of the hyperparameters and the corresponding network parameters within a pre‐defined hyperparameter search space. A total of 16‐year data from Millstone Hill incoherent scatter radar (ISR) are used for the NN models. One single‐layer NN (SLNN) model and one deep NN (DNN) model are both trained with NAS, namely SLNN‐NAS and DNN‐NAS, forNeprediction and compared with their manually tuned counterparts (SLNN and DNN) based on previous studies. Our results show that SLNN‐NAS and DNN‐NAS outperformed SLNN and DNN, respectively. These NN predictions ofNedaily variation patterns reveal a 27‐day mid‐latitude topsideNevariation, which cannot be reasonably represented by traditional empirical models developed using monthly averages. DNN‐NAS yields the best prediction accuracy measured by quantitative metrics and rankings of daily pattern prediction, especially with an improvement in mean absolute error more than 10% compared to the SLNN model. The limited improvement of NAS is likely due to the network complexity and the limitation of fully connected NN without the time histories of input parameters. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email