Temperature Fusion of TIMED/SABER Data and COSMIC Data in Stratosphere.

The acquisition of accurate stratospheric temperature data is of great significance for the research on global climate system and numerical weather forecasting. To obtain high‐precision stratospheric temperature, the three‐dimensional variational assimilation method, combined with kriging interpolation, was applied for data fusion of the Thermosphere, Ionosphere, Mesosphere Energetics, and Dynamics/Sounding of the Atmosphere using Broadband Emission Radiometry (TIMED/SABER) temperature data and the Constellation Observing System for Meteorology, Ionosphere and Climate (COSMIC) radio occultation temperature data. As a sample, the two data sets of TIMED/SABER and COSMIC, during the time range from 5 December 2009, 09:00 to 5 December 2009, 15:00, were used to construct global stratospheric temperature field of 5 December 2009, 12:00. The cross‐validation results showed that temperature distribution from the variational fusion method was consistent with COSMIC observations at the height range of 15–35 km, which the mean deviations and standard deviations were within a range of −1 to 1 and 3 to 4 K, respectively. At the height range from 35 to 40 km, the temperature distribution from the variational fusion method was more consistent with TIMED/SABER observations. The mean deviations were within 0 and 0.5 K, and the standard deviations were between 1.8 and 2 K. Overall, the temperature field after variational fusion combines the high‐precision advantage of SABER observations around 40 km and the high‐precision advantage of COSMIC observations at the height range of 15–30 km, the effect of the variational fusion method is significant. Key Points: The three‐dimensional variational assimilation method was applied for the data fusion of Thermosphere, Ionosphere, Mesosphere Energetics, and Dynamics/Sounding of the Atmosphere using Broadband Emission Radiometry data and Constellation Observing System for Meteorology, Ionosphere and Climate data in stratosphereThe mean deviations and the standard deviations between the variational fusion data and observations were small in stratosphereThe effect of the variational fusion method was significant [ABSTRACT FROM AUTHOR]