Evaluation of the New Background Covariance Model for the Ionospheric Data Assimilation

Forsythe, Victoriya V.; Azeem, Irfan; Blay, Ryan; Crowley, G.; Gasperini, Federico; Hughes, J. D.; Makarevich, R. A.; Wu, Wanli

doi:10.1029/2021rs007286

Cited by 17 publications

(25 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Forsythe et al. (2020, 2021) constructed global maps of ionospheric correlation lengths and developed new background covariance models for ionospheric data assimilation. These avenues provide a useful step toward improving the specification of background error covariance. Devising high‐precision regional data assimilation schemes to specify detailed localized ionospheric weather features: Most current ionospheric data assimilation models have been built to run on a global scale that may not always have optimal performance in representing local mesoscale ionosphere morphology.…”

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

“…Recently, some studies used hybrid assimilation approaches to combine the merits of variational methods and ensemble-based algorithms (e.g., Schwartz et al, 2014;. Forsythe et al (2020Forsythe et al ( , 2021 constructed global maps of ionospheric correlation lengths and developed new background covariance models for ionospheric data assimilation. These avenues provide a useful step toward improving the specification of background error covariance.…”

mentioning

confidence: 99%

See 2 more Smart Citations

3‐D Regional Ionosphere Imaging and SED Reconstruction With a New TEC‐Based Ionospheric Data Assimilation System (TIDAS)

Zhang

Erickson

et al. 2022

Space Weather

View full text Add to dashboard Cite

A new TEC‐based ionospheric data assimilation system (TIDAS) over the continental US and adjacent area (20°–60°N, 60°–130°W, and 100–600 km) has been developed through assimilating heterogeneous ionospheric data, including dense ground‐based Global Navigation Satellite System (GNSS) Total Electron Content (TEC) from 2,000+ receivers, Constellation Observing System for Meteorology, Ionosphere, and Climate radio occultation data, JASON satellite altimeter TEC, and Millstone Hill incoherent scatter radar measurements. A hybrid Ensemble‐Variational scheme is utilized to reconstruct the regional 3‐D electron density distribution: a more realistic and location‐dependent background error covariance matrix is calculated from an ensemble of corrected NeQuick outputs, and a three‐dimensional variational (3DVAR) method is adopted for measurement updates to obtain an optimal state estimation. The spatial‐temporal resolution of the reanalyzed 3‐D electron density product is as high as 1° × 1° in latitude and longitude, 20 km in altitude, and 5 min in universal time, which is sufficient to reproduce ionospheric fine structure and storm‐time disturbances. The accuracy and reliability of data assimilation results are validated using ionosonde and other measurements. TIDAS reanalyzed electron density is able to successfully reconstruct the 3‐D morphology and dynamic evolution of the storm‐enhanced density (SED) plume observed during the St. Patrick's day geomagnetic storm on 17 March 2013 with high fidelity. Using TIDAS, we found that the 3‐D SED plume manifests as a ridge‐like high‐density channel that predominantly occurred between 300 and 500 km during 19:00–21:00 UT for this event, with the F2 region peak height being raised by 40–60 km and peak density enhancement of 30%–50%.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

3‐D Regional Ionosphere Imaging and SED Reconstruction With a New TEC‐Based Ionospheric Data Assimilation System (TIDAS)

Zhang

Erickson

et al. 2022

Space Weather

View full text Add to dashboard Cite

show abstract

“…Prior work from (Forsythe, Azeem, Blay, et al, 2021) found similar structure in the vertical correlation length in that the correlation length increases with altitude and is typically between 50 and 200 km, however they do not use a 200 km limit so their high altitude correlation lengths are much larger. Our use of exclusively bottomside ionosonde data also contributes to this difference.…”

Section: Vertical Correlationmentioning

confidence: 81%

On Constructing a Realistic Truth Model Using Ionosonde Data for Observation System Simulation Experiments

Hughes¹,

Forsythe

Blay³

et al. 2022

Radio Science

Self Cite

View full text Add to dashboard Cite

The ionosphere contains many small‐scale electron density variations that are under represented in smooth physics‐based or climatological models. This can negatively impact the results of Observation System Simulation Experiments, which use a truth model to simulate data. This paper addresses this problem by using ionosonde data to study ionospheric variability and build a new truth model with empirically driven variations. The variations are studied for their amplitude, horizontal and vertical size, and temporal extent. Results are presented for different local times, seasons, and at solar minimum and solar maximum. We find that these departures from a smooth background are often as large as 25% and are most prevalent near 250 km in altitude. They have horizontal spatial extents that vary from a few hundred to a few thousand kilometers, and typically have the largest horizontal extent at high altitudes. Their vertical extents follow the same pattern of being larger at high altitudes, but they only vary from 10s of km up to 200 km in vertical size. Temporally, these variations can last for a few hours. A procedure for using these spatial and temporal distributions to add empirically driven variance to a smooth truth model is outlined. This process is used to make a truth model with representative variations, which is compared to ionosonde data as well as Global Positioning System Total Electron Content data that was not used to inform the model. The new model resembles the data much better than the smooth models traditionally used.

show abstract

“…This is because of the influence of horizontal correlation lengths from nearby GPS stations over southern Europe (Figures 2a–2d). In fact, IDA4D NmF2 results over mid‐latitude are more accurate than IRI NmF2 (V. V. Forsythe et al., 2021; Jeong et al., 2022; Mengist et al., 2019). The accuracy of IDA4D F2 peak density altitude (hmF2) is not discussed in this study because assimilating GPS‐STEC or GPS‐STEC and C2 NmF2 (Figure 4) has little or no effect on IRI hmF2 improvement (Ssessanga et al., 2019).…”

Section: Results and Validationmentioning

confidence: 99%

3‐D Regional Imaging of Ionosphere Over Africa Through Assimilating Satellite and Ground‐Based Data

et al. 2023

View full text Add to dashboard Cite

In this study, the first high‐resolution regional ionospheric model over Africa and adjacent areas (−40–40°N latitude, 30°W–60°E longitude, and 80–1,400 km in altitude) is constructed by assimilating ground‐based slant total electron content (STEC) from 40 GPS (Global Positioning System) receiver stations and space‐based NmF2 (ionospheric F2 peak density) data from C2 (Constellation Observing System for Meteorology, Ionosphere, and Climate‐2) into the International Reference Ionosphere (IRI‐2016) model. An Ionospheric Data Assimilation Four‐Dimensional (IDA4D) technique was used to estimate electron densities as high as 1.5° × $\times $ 3° in latitude and longitude, 10 km in altitude in the E and F regions, and 15 min in universal time. Two experiments were run for the following data sets: (a) GPS‐STECs only and (b) GPS‐STECs and NmF2s from C2 during geomagnetically quiet (6–11 May 2021) and storm periods (12–14 May 2021). The IDA4D assimilation results are validated using independent C2 control group, ionosonde, and JASON‐3 observations. Results for the storm period show that experiment 2 reduces the average root‐mean‐square error (RMSE) of NmF2, foF2, and VTEC by 34%, 31%, and 34%, respectively, and increases the associated correlations by 10%, 14%, and 2% over IRI, respectively. Using IDA4D, we observed enhancement of the northern crest equatorial ionization anomaly in the late evening that was caused by upward and northward plasma transport.

show abstract

Evaluation of the New Background Covariance Model for the Ionospheric Data Assimilation

Cited by 17 publications

References 17 publications

3‐D Regional Ionosphere Imaging and SED Reconstruction With a New TEC‐Based Ionospheric Data Assimilation System (TIDAS)

3‐D Regional Ionosphere Imaging and SED Reconstruction With a New TEC‐Based Ionospheric Data Assimilation System (TIDAS)

On Constructing a Realistic Truth Model Using Ionosonde Data for Observation System Simulation Experiments

3‐D Regional Imaging of Ionosphere Over Africa Through Assimilating Satellite and Ground‐Based Data

Contact Info

Product

Resources

About