Highlights of CRTM Development and Validation

1. Highlights of CRTM Development and Validation By Yong Chen Thanks: F. Weng, Y. Han and CRTM team Presented to CIRA Director, Professor Chris Kummerow Highlights, Y. Chen

2. Requirements on CRTM • Perform fast and accurate forward, tangent linear/adjoint calculations • Support all the satellite instruments (US and foreign) that are used in NWP models • Work under all atmospheric and surface conditions • Have a flexible interface with different NWP models such as GFS, NOGAPS, and WRF • Allow future expansion for broader applications • CRTM supports more than 100 Sensors • GOES-R ABI • NPP/JPSS CrIS/ATMS/VIIRS • MetopA and B IASI/HIRS/AVHRR/AMSU/MHS • TIROS-N to NOAA-19 AVHRR • TIROS-N to NOAA-19 HIRS • GOES-8 to 15 Imager • GOES-8 to 15 sounder IR and VIS • Terra/Aqua MODIS • MSG SEVIRI • Aqua AIRS, AMSR-E, AMSU-A,HSB • NOAA-15 to 19 AMSU-A • NOAA-15 to 17 AMSU-B • NOAA-18/19 MHS • TIROS-N to NOAA-14 MSU • DMSP F8, F10, F11, and 13 to15 SSM/I • DMSP F13,15 SSM/T1 • DMSP F14,15 SSM/T2 • DMSP F16-20 SSMIS • CoriolisWindsat • TiROS-NOAA-14 SSU • FY-3 A and B IRAS, MERSI, MWTS,MWHS,MWRI • COMS MI • GPM GMI • TRMM TMI • MT MADRAS/SAPHIR Highlights, Y. Chen

3. CRTM Major Modules public interfaces Forward model Tangent-linear model Adjoint model Jacobian model CRTM Clearance CRTM Initialization SfcOptics (Surface Emissivity Reflectivity Models) AerosolScatter (Aerosol Absorption Scattering Model) CloudScatter (Cloud Absorption Scattering Model) Moleculescatter (Molecular scattering model) AtmAbsorption (Gaseous Absorption Model) RTSolution (RT Solver) Source Functions Highlights, Y. Chen

4. General Transmittance Models Two general atmospheric transmittance models implemented: • ODAS (Optical depth in absorber space) - OPTRAN • Optical depth computed in the coordinates of integrated absorber amount • Variable gases: H2O and O3 • ODPS (Optical Depth in Pressure Space) • Optical depth computed in pressure coordinates • Variable absorbing gases: H2O, CO2, O3, CO, N2O and CH4 • Water vapor line computed using ODAS (optional) CRTM simulated brightness temperature spectra for hyper-spectral infrared sensors IASI (black), AIRS (red) and CrIS (blue). Highlights, Y. Chen

5. General Transmittance Models • In the ODPS transmittance model, the ODAS (OPTRAN) algorithm is used (optional) to compute water vapor line transmittances since it can provide better forward results and Jacobians for many IR channels. • Results have shown positive NWP impacts from temperature related fields when the ODPS plus ODAS for H2O line is used, compared to the use of ODPS alone. Temperature Jacobians Water vapor Jacobians Chen et al., JGR, 2010 Highlights, Y. Chen

6. Fast Transmittance Model for Stratospheric Sounding Unit (SSU) • The SSU channel spectral response function (SRF) is a combination of the instrument filter function and the transmittance of a CO2 cell. • The SRF varies due to the cell CO2 leaking problem. • CRTM-v2 includes schemes to take the SRF variations into account (Liu and Weng, 2009; Chen et al., JTECH 2011). • CO2 and O3 are variables gases • The SSU model can be applied for reanalysis of the observations, also be used to address two important corrections in deriving trends from SSU measurements: CO2 cell leaking correction, and atmospheric CO2 concentration correction. CRTM simulations compared with SSU observations for SSU noaa-14. CO2 cell pressure variations, which causes SSU SRF variations. Highlights, Y. Chen

7. Multiple Transmittance Algorithms Framework CRTM Initialization: load transmittance coefficient data User specified sensor ID array TauCoeff files Algorithm ID Sensor ID Coeff. Data Load TauCoeff. Data Memory CRTM transmittance models: ODAS, ODPS, ODSSU, ODZeeman Algorithm selection (Algorithm ID) ODAS • • • • • • ODPS Optical depth profile Other modules Highlights, Y. Chen

8. Comparison of ODAS and ODPSApplication to AVHRR • The channel 3 bias shows a strong zenith angle dependence when the zenith angle is larger than 40o. This bias is due probably to the cloud contamination at larger zenith angles, and additional absorption in the atmosphere associated with aerosol and other greenhouse gases such as carbon dioxide and water • In the ODPS algorithm, the reducing radiance due to CFCs absorption is partially offset by the increasing radiance due to smaller local zenith angle at higher atmospheric levels. • Overall, ODPS is better than ODAS. The biases between simulations and observations at channel 4 and 5 are dramatically improved in ODPS and very close to zero. The inclusion of CFCs in ODPS reduces bias by 0.25 K at channel 4 and 0.15 K at channel 5 at nadir. Chen et. al., JGR, 2011, submitted Highlights, Y. Chen

9. Histograms of the Observed and Simulated for AMSUA, MHS BTs over Ocean Observation Simulation • Coincidental/collocated NOAA 18 AMSUA/MHS satellite observations and CloudSat cloud profiles (IWC, LWC) and ECMWF analysis atmospheric profiles as input in CRTM. • Reasonable agreements of observed and simulated BT distributions at all frequencies. Chen et al., JGR, 2008 Highlights, Y. Chen

10. Histograms of the BT Difference (O–S) over Ocean Cloudy Clear • The distributions are in Gaussian shapes with maximum observation at or near zero, which confirm that the agreement between observed and simulated BTs are very good under clear and cloudy conditions. • There are clear-sky biases in certain surface sensitive microwave channels of the order of 1–2 K which is due to the sea-surface emission model used in CRTM. Highlights, Y. Chen

11. Ongoing and Future Work • Test of new microwave land emissivity database in CRTM and study of its impacts in GDAS. • New transmittance coefficients for future satellite sensors (GMI, FY-3C/D etc). • Validation of CRTM in cloudy and rainy conditions using aircraft observations. An aircraft version of CRTM has been implemented. Highlights, Y. Chen