An official website of the United States government
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
Daily Predictions of F10.7 and F30 Solar Indices With Deep Learning
Abstract The F10.7 and F30 solar indices are the solar radio fluxes measured at wavelengths of 10.7 and 30 cm, respectively, which are key indicators of solar activity. F10.7 is valuable for explaining the impact of solar ultraviolet (UV) radiation on the upper atmosphere of Earth, while F30 is more sensitive and could improve the reaction of thermospheric density to solar stimulation. In this study, we present a new deep learning model, named the Solar Index Network, or SINet for short, to predict daily values of the F10.7 and F30 solar indices. The SINet model is designed to make medium‐term predictions of the index values (1–60 days in advance). The observed data used for SINet training were taken from the National Oceanic and Atmospheric Administration as well as Toyokawa and Nobeyama facilities. Our experimental results show that SINet performs better than five closely related statistical and deep learning methods for the prediction of F10.7. Furthermore, to our knowledge, this is the first time deep learning has been used to predict the F30 solar index.
more »
« less
- PAR ID:
- 10674739
- Publisher / Repository:
- Wiley
- Date Published:
- Journal Name:
- Journal of Geophysical Research: Space Physics
- Volume:
- 131
- Issue:
- 2
- ISSN:
- 2169-9380
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Ground-based temperature measurements at Svalbard, Wuppertal, and Hohenpeissenberg were analyzed to obtain F10.7, Ap index, and Dst index trends. The trends were then compared to those obtained from Sounding of the Atmosphere using Broadband Emission Radiometry (SABER) temperature measurements at the same locations. Trend analysis was carried out for overlapped time periods, full range of available data, and the CO2-detrended full range of available data. The Svalbard meteor radar (SABER) temperature showed a weak (moderate) correlation with F10.7 and a moderate (weak) correlation with Ap and Dst indices. The trends in the Wuppertal OH* temperature compare well with the SABER temperature when a full range of data is used in the analysis. Both temperatures had a similar F10.7 trend with the same level of correlation coefficient. The F10.7 trend in the Hohenpeissenberg OH* temperature compared well with that obtained by SABER, but the former displayed a weak correlation. The Hohenpeissenberg data displayed a very weak correlation with Ap and Dst indices. Our study clearly shows that a longer dataset would better capture trends in temperature, as was evidenced by the results of Wuppertal data. The CO2-detrended temperatures overall showed slightly larger trend values with a slightly better correlation.more » « less
-
Abstract EUropean Heliospheric FORecasting Information Asset (EUHFORIA) is a physics‐based data‐driven solar wind and coronal mass ejections (CMEs) propagation model designed for space weather forecasting and event analysis investigations. Although EUHFORIA can predict the solar wind plasma and magnetic field properties at Earth, it is not equipped to quantify the geo‐effectiveness of the solar transients in terms of geomagnetic indices like the disturbance storm time (Dst) index and the auroral indices, that quantify the impact of the magnetized plasma encounters on Earth's magnetosphere. Therefore, we couple EUHFORIA with the Open Geospace General Circulation Model (OpenGGCM), a magnetohydrodynamic model of the response of Earth's magnetosphere, ionosphere, and thermosphere to transient solar wind characteristics. In this coupling, OpenGGCM is driven by the solar wind and interplanetary magnetic field obtained from EUHFORIA simulations to produce the magnetospheric and ionospheric response to the CMEs. This coupling is validated with two observed geo‐effective CME events driven with the spheromak flux‐rope CME model. We compare these simulation results with the indices obtained from OpenGGCM simulations driven by the measured solar wind data from spacecraft. We further employ the dynamic time warping (DTW) technique to assess the model performance in predicting Dst. The main highlight of this study is to use EUHFORIA simulated time series to predict the Dst and auroral indices 1–2 days in advance, as compared to using the observed solar wind data at L1, which only provides predictions 1–2 hr before the actual impact.more » « less
-
Abstract We propose a novel deep learning framework, named SYMHnet, which employs a graph neural network and a bidirectional long short‐term memory network to cooperatively learn patterns from solar wind and interplanetary magnetic field parameters for short‐term forecasts of the SYM‐H index based on 1‐ and 5‐min resolution data. SYMHnet takes, as input, the time series of the parameters' values provided by NASA's Space Science Data Coordinated Archive and predicts, as output, the SYM‐H index value at time pointt + whours for a given time pointtwherewis 1 or 2. By incorporating Bayesian inference into the learning framework, SYMHnet can quantify both aleatoric (data) uncertainty and epistemic (model) uncertainty when predicting future SYM‐H indices. Experimental results show that SYMHnet works well at quiet time and storm time, for both 1‐ and 5‐min resolution data. The results also show that SYMHnet generally performs better than related machine learning methods. For example, SYMHnet achieves a forecast skill score (FSS) of 0.343 compared to the FSS of 0.074 of a recent gradient boosting machine (GBM) method when predicting SYM‐H indices (1 hr in advance) in a large storm (SYM‐H = −393 nT) using 5‐min resolution data. When predicting the SYM‐H indices (2 hr in advance) in the large storm, SYMHnet achieves an FSS of 0.553 compared to the FSS of 0.087 of the GBM method. In addition, SYMHnet can provide results for both data and model uncertainty quantification, whereas the related methods cannot.more » « less
-
Abstract We develop an Imbalanced Regression Artificial Neural Network model for the Auroral electrojet index (IRANNA) to predict the SuperMAG SML index, addressing the heavily imbalanced distribution of the SML data set. The data set contains mostly quiet‐time values of lesser importance and very few strong‐to‐extreme values of interest, such as those associated with super substorms. Traditional prediction models, which minimize mean squared error uniformly across the whole data set, are often skewed by this imbalance, prioritizing the lower, quiet‐time values and consequently underestimating strong geomagnetic events. The IRANNA model addresses this issue by using a customized weighting scheme in the loss function, enabling it to predict strong‐to‐extreme events accurately for the first time. The model takes solar wind parameters as inputs and predicts the logarithm of the absolute SML values. It does not rely on past values of the SML index, differentiating it from other models that use historical data for prediction. The model has demonstrated its ability to predict the peak amplitudes of strong‐to‐extreme events across various statistical analyses, event studies, and virtual experiments. Despite this success, challenges remain, particularly during localized electrojet events and when upstream solar wind data propagation is unreliable. This study emphasizes the importance of using imbalanced regression techniques, especially in space physics, where data sets are inherently skewed. It also highlights the potential of the IRANNA model to provide valuable insights into the magnetosphere's response to solar wind driving, improving space weather forecasting and offering new tools for investigating magnetospheric dynamics.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript
- Journal Article:
- https://doi.org/10.1029/2025JA034868
