Combining radio occultation (RO) and IR/MW measurements to infer temperature and moisture profiles

Participants: Eva Borbas , Paul Menzel, Jun Li

Method

The NOAA88 radiosonde dataset was used to simulate IR and MW brightness temperatures, GPS refractivities and surface observations. The IR data represent the HIRS (ATOVS) on polar orbiting NOAA satellites (also represent the GOES Sounder) and the Crosstrack Infrared Sounder (CrIS), the MW sounder data represent the AMSU (ATOVS) instrument.

The statistical regression approach was used for getting temperature and humidity profiles. Both linear and quadratic terms for brightness temperatures and refractivities (HIRS, CrIS, AMSU, GPS, and surface observations) were included in the regression relationship. Regression coefficients were derived from 90 % of all profiles and tested in the remaining 10 %.

The impact of HIRS, CrIS, AMSU, GPS, and SFC data on the temperature and humidity retrievals was studied. The rms errors of the retrievals obtained from the different combinations of information were compared to the original NOAA88 data.

Figure

The figure shows the difference of various rms errors for the temperature and humidity retrievals. The improvements of GPS information on temperature (first column) and on humidity (second column) retrievals without AMSU (upper panels) and with (lower panels) are shown. The upper right panel shows the impact of AMSU. The lower right panel shows the bias and rms errors of the “best “ case using ATOVS (AMSU+HIRS) and GPS or (CrIS+AMSU) + GPS. The red solid lines stand for the case when the CrIS was used as IR information and black dashed line when IR information was the HIRS data.

The Figure show the GPS information improves the HIRS and ATOVS temperature and humidity retrievals, and it also can improve the temperature retrievals derived from the CrIS 393 super channels around and above the tropopause.

Or the GPS either the AMSU information has very slight effect on humidity retrievals obtained from CrIS data.

Summary

In this study of simulated retrievals performed using a statistical regression approach, the combination of radiometric (IR and MW) and geometric (RO) information yields improved tropospheric temperature and moisture profiles when compared to those inferred from either system alone.

Specifically in the troposphere GPS improves:

·  HIRS (representative of GOES Sounder) temperature profile retrievals from the tropopause by 1.5 C down to 570 hPa by 0.25 K.

·  ATOVS (AMSU plus HIRS) temperature profile retrievals around the tropopause level by 0.5K.

·  CrIS temperature profile retrievals around the tropopause level by 0.3 K

· 

•  HIRS (representative of GOES Sounder) moisture profile retrievals from 400 hPa by 3 % to 700 hPa by about 5 %.

•  ATOVS moisture profile retrievals from 570 to 700 by about 4 %.

·  Doubling GPS refractivity noise degrades the ATOVS temperature profile improvement by about 0.2 K from 25 to 350 hPa and humidity profile improvement by 1 % from 570 to 700 hPa.

In the stratosphere GPS improves

•  ATOVS temperature profile retrievals between 15 hPa and the tropopause by about 0.5K.

•  CrIS temperature profile retrievals between 10 hPa and the tropopause by about 0.4K