The YUNYAO Global Navigation Satellite System Radio Occultation (GNSS RO) products are an important component of China’s commercial radio occultation datasets. With the rapid development of commercial low-Earth-orbit satellite constellations, it is necessary to assess the quality of these new GNSS RO products using mature satellite missions and independent ground-based observations. In this study, YUNYAO ionospheric products from February-March and July 2025 and YUNYAO neutral atmospheric profile products from July 2025 are evaluated. COSMIC-2 RO products, Global Ionosphere Radio Observatory (GIRO) ionosonde data, and Integrated Global Radiosonde Archive (IGRA) radiosonde data are used as reference datasets. The ionospheric evaluation focuses on the F2-layer peak electron density (NmF2) and peak height (hmF2) extracted from ionPrf electron density profiles. The neutral atmospheric evaluation includes temperature, refractivity, water vapor pressure, and specific humidity profiles.

For the ionospheric products, profile quality control and spatiotemporal matching were first performed before statistical comparison. The YUNYAO-COSMIC-2 comparison shows good consistency in both F2-layer peak parameters. The correlation coefficients of NmF2 and hmF2 are 0.932 and 0.956, respectively, and the mean biases are −7.37×10⁹ el·m⁻³ and 0.62 km, respectively. These results indicate that YUNYAO and COSMIC-2 have similar capability in describing the overall variation of F2-layer peak parameters, especially for hmF2. The comparison between YUNYAO and GIRO ionosonde data also shows a clear positive correlation. The correlation coefficients of NmF2 and hmF2 are 0.934 and 0.739, respectively, with mean biases of 1.12×10¹¹ el·m⁻³ and 11.80 km, respectively. Compared with the YUNYAO-COSMIC-2 comparison, the YUNYAO-GIRO comparison shows larger bias and dispersion, which may be related to differences in observing geometry, spatial representativeness, local ionospheric horizontal gradients, and the automatic interpretation uncertainty of ionosonde parameters.

For the neutral atmospheric products, YUNYAO wetPrf profiles were compared with COSMIC-2 wetPf2 products and IGRA radiosonde profiles after interpolation to standard pressure levels. Temperature and refractivity were evaluated within 925-10 hPa, while water vapor pressure and specific humidity were evaluated within 925-200 hPa because water vapor content is very low in the upper levels. Compared with COSMIC-2, the mean temperature bias and mean relative refractivity bias of YUNYAO are −0.025 K and 0.05%, respectively, within 925-10 hPa. Compared with IGRA, the corresponding values are −0.028 K and 0.12%, respectively. These results suggest that YUNYAO has good consistency with both RO and radiosonde reference datasets in temperature and refractivity. For humidity-related variables, the mean relative biases of water vapor pressure between YUNYAO and COSMIC-2 and between YUNYAO and IGRA are 18.23% and 18.24%, respectively, within 925-200 hPa. The corresponding mean specific humidity biases are 0.010 g/kg and -0.092 g/kg, respectively. The specific humidity biases are small in magnitude, indicating that YUNYAO and the reference datasets are generally close in water vapor content, while the relative bias of water vapor pressure shows more evident variations among different pressure levels. This difference is likely associated with the enhanced sensitivity of relative water vapor pressure statistics to small background values under low-humidity conditions.

Overall, the YUNYAO GNSS RO products show good data quality in F2-layer ionospheric peak parameters, as well as in neutral atmospheric temperature, refractivity, and humidity profiles. The agreement with COSMIC-2, GIRO, and IGRA indicates that YUNYAO products have potential value for ionospheric monitoring and atmospheric vertical structure analysis. Meanwhile, the interpretation of humidity-related variables should consider the influence of low-humidity background conditions, retrieval methods, and reference-data uncertainties. The results of this study can provide useful data references for ionospheric monitoring, atmospheric structure analysis, and numerical weather prediction data assimilation. Further assessments based on longer time series and more diverse seasonal, local-time, latitudinal, and space-weather conditions would help examine the multi-scenario applicability of YUNYAO GNSS RO products.