Technical Article

Estimating COSMIC ionospheric radio occultation foF₂ error using the three-cornered hat method

¹ Space Weather Technology, Research and Education Center (SWx TREC), University of Colorado Boulder, 80309 Boulder, CO, USA
² State Key Laboratory of Space Weather, National Space Science Center, Chinese Academy of Sciences, 100190 Beijing, PR China
³ University of Chinese Academy of Sciences, 101408 Beijing, PR China
⁴ Key Laboratory of Science and Technology on Environmental Space Situation Awareness, Chinese Academy of Sciences, 100190 Beijing, PR China

^* Corresponding author: xcwu@nssc.ac.cn

Received: 1 April 2025
Accepted: 28 November 2025

Abstract

GNSS ionospheric radio occultation (IRO) is an important technique for remote sensing of Earth’s ionosphere. Previous validation of GNSS IRO has often been based on comparisons with ground-based observations, such as ionosondes or in-situ observations, which are assumed to be true representations of the real ionospheric state. But these reference observations commonly used for validation also have errors of their own, which are often neglected in such comparisons. In this study, we employ the three-cornered hat (3CH) method to analyze error statistics and separate errors present in Constellation Observing System for Meteorology, Ionosphere, and Climate (COSMIC) IRO foF₂ observations. This provides a better characterization of the errors since it avoids the need to know the true value of the quantity of interest. To form the necessary triplet datasets, ionosondes, International Reference Ionosphere (IRI), NeQuick, and the Whole Atmosphere Community Climate Model with thermosphere and ionosphere extension (WACCM-X) data are used in addition to COSMIC. To verify the feasibility of this method under the error conditions of the data used in this paper, we initially verified the method using simulated data. The error characteristics are analyzed according to local time and geomagnetic latitude. IRO electron density error characteristics are found to be primarily related to inversion assumptions. The COSMIC IRO errors are smallest in summer and largest in winter, and the errors in high solar activity years are larger than the errors in low solar activity years. When understanding error results derived from the 3CH method, special attention should be given to the impact of error correlation and representativeness errors. The results show that the 3CH method provides a better characterization of the errors in COSMIC IRO observations, is helpful for the specification of errors in data assimilation systems, and is suitable for estimating errors in geospace observations.