Evaluation of the retrieved temperature and specific humidity from COSMIC-2 and FY-3D with Radiosonde, reanalysis data, and MetOp

Abstract

The Constellation Observing System for Meteorology, Ionosphere, and Climate-2 (COSMIC-2), as the follow-on mission to COSMIC, commenced observations from 25 June 2019. COSMIC-2, with its high signal-to-noise ratio (SNR) and excellent stability, is expected to outperform COSMIC in detecting atmospheric parameters while penetrating the lower atmosphere.Therefore, to verify and analyze the detection capability of COSMIC-2 occultation in the neutral atmosphere, we retrieved temperature and specific humidity from COSMIC-2 and FY-3D based on the Able integral and the one-dimensional variational method. The evaluations are carried out on retrieval parameters at different altitudes using official RO products, radiosonde, reanalysis data, and MetOp. A detailed analysis of errors in the evaluation process is provided. The RMSEs for COSMIC-2 temperature and specific humidity are less than 1.5 K and 36 % respectively when radiosonde as validation data. Compared to FY-3D, accuracy improved by 11.1 % and 18.5 % respectively. FY-3D shows better consistency in the evaluation of the RO official products and radiosonde. Temporal and spatial analysis reveals minimal seasonal differences but significant latitude influence on evaluation outcomes. In the evaluation of reanalysis data, COSMIC-2 temperature and NCEP performed well in the troposphere, with a correlation coefficient greater than 0.9. The RMSE of the COSMIC-2 temperature and specific humidity in the MERRA-2 evaluation increased by 50.59 % and 16.65 % over FY-3D. Besides, derived parameters are more consistent with MetOp than radiosonde. Moreover, the co-localization threshold and resampling have less than 0.1 K and 5 % impact on the accuracy of evaluating temperature and specific humidity.