Premium
An Evaluation of Spire Radio Occultation Data in Assimilative Ionospheric Model GPSII and Validation by Ionosonde Measurements
Author(s) -
Kramer Kelsey K.,
Fridman Sergey V.,
Nickisch L. J.
Publication year - 2021
Publication title -
radio science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.371
H-Index - 84
eISSN - 1944-799X
pISSN - 0048-6604
DOI - 10.1029/2020rs007139
Subject(s) - ionosonde , gnss applications , radio occultation , ionosphere , standard deviation , geodesy , occultation , international reference ionosphere , remote sensing , satellite , total electron content , ionogram , geology , meteorology , physics , mathematics , tec , geophysics , statistics , electron density , quantum mechanics , astronomy , electron
This paper presents a comprehensive analysis of results obtained using NWRA’s assimilative ionospheric model GPSII when it is driven by Spire Global, Inc.’s global navigation satellite system ‐ radio occultation (GNSS‐RO) measurements. Results of the assimilative model are validated using the database of Digisonde and Dynasonde ionograms. A week’s worth of dual‐frequency phase data was collected from Spire’s constellation of low‐Earth orbiting micro‐satellites carrying GNSS receivers and assimilated into GPSII. A validation study of GNSS‐RO measurements was performed in four geographical regions. GPSII operation with the data was evaluated and the results were validated against independent ionospheric measurements. Ground‐based ionosondes were employed in diagnosing the accuracy of the derived models; we studied the differences between the F2 peak parameters (the F2 layer critical frequency,f 0 F 2 , and the peak height of the F2 layer,H m F 2 ) derived from ionograms and those computed with GPSII. For the 1,491 validation events used in this study, the magnitude of the deviation of the GPSII‐predictedf 0 F 2at an ionosonde location from the corresponding ionosonde‐observed value was found on average to be 0.608 MHz (with a median of 0.461 MHz) and the magnitude of the deviation of the GPSII‐predictedH m F 2 at an ionosonde location from the corresponding ionosonde‐observed value was found to be on average 22.9 km (with a median value of 18.0 km). Median GPSII validation errors forf 0 F 2are a factor of four smaller than corresponding errors for IRI, while GPSIIH m F 2errors are a factors of two smaller than those for IRI. The analysis showed that Spire’s TEC data was both compatible for ingestion into GPSII and that the assimilation of the data improved the comparison to ground‐based ionosondes.