4.2 Improving the temperature and water vapor retrievals in the upper troposphere and stratosphere using combined AIRS, AMSU, and GPS RO measurements

In this section, we used GPS RO data to constrain the AIRS and AMSU temperature retrievals serving to improve moisture retrievals in the upper troposphere and lower stratosphere (UT/LS). In the upper troposphere, GPS RO refractivity is very sensitive to temperature but less sensitive to moisture. It is demonstrated by Ho et al. [34] that GPS RO refractivity can resolve temperatures greater than 1 K around 200 hPa but it can only sense about 15% of water vapor variation. Figure 7 shows the temperature and moisture retrieval RMSE for AIRS, AMSU, and GPS RO as well as the combined AIRS, AMSU, and GPS RO data. The multi-variable

#### Figure 7.

The RMSE of temperature for AIRS, AMSU, GPS, and AIRS+AMSU+GPS in the UT/LS is on the left panel, and the RMSE of water vapor for AIRS, AMSU, GPS, and AIRS+AMSU+GPS in the UT/LS is on the right panel.

regression method described in Section 4.1 is used for the data from the simulation study. The 100 level AIRS vertical grids are used for all AIRS, AMSU, and GPS RO data. With a very high vertical resolution GPS RO refractivity profiles, AIRS and AMSU temperature RMSEs improved from 0.8 and 1.0 K, respectively, between 250 and 100 hPa (tropopause layer) to 0.4 K, which lead to AIRS and AMSU moisture RMSE around the same layer decrease from 4 and 15 ppmv, respectively, to around than 3 ppmv.

Since the open-loop tracking algorithm is only applied to COSMIC data, GPS RO data from COSMIC are used with AIRS data to derive moisture and temperature profiles in the clear skies of the free troposphere. For the UT/LS retrievals, GPS RO from SAC-C, GRACE, CHAMP, COSMIC, and GRAS data can be used to collocate with AIRS data. It can be seen in the left panel of Figure 6 that the region of the largest temperature gradient is around 200 hPa, where the temperature RMSE for AIRS and AMSU is around 1.0 K. The fact that much improved temperature retrievals from GPS RO data (RMSE is 0.6 K) and from the combined AIRS, AMSU, and GPS RO data (RMSE is 0.3 K) are very useful to construct accurate temperature and moisture structures in the UT/LS region for the entire globe.
