Simulating the IASI Spectrum using Radiative Transfer Models
Ventress, Lucy; Dudhia, Anu
AOPP, University of Oxford, UNITED KINGDOM

The Infrared Atmospheric Sounding Interferometer (IASI) is a nadir viewing infrared Fourier transform spectrometer onboard the METOP-A satellite. Covering the spectral range 645-2760 cm -1 with a spectral sampling of 0.25 cm-1 , results in a large quantity of data. This has allowed it to become a key instrument in providing temperature and water vapour profiles and its data have been extensively assimilated in Numerical Weather Prediction (NWP) systems.
The ability to accurately simulate the IASI spectra using Radiative Transfer Models is essential within the data assimilation process and in continuing to produce accurate retrievals. Here we discuss comparisons between two Radiative Transfer Models; the Reference Forward Model (RFM), a line by line radiative transfer model developed at the University of Oxford, and the Radiative Transfer Model for TOVS (RTTOV), a widely used fast radiative transfer model adapted for IASI. A comparison between the models is carried out to establish the origin of spectral discrepancies between them.
The RFM performs accurate radiance calculations at high spectral resolution as an integral over both atmospheric path and wavenumber, whereas RTTOV, in order to increase computer efficiency, calculates spectrally averaged transmittances integrated over the atmospheric path only. The RTTOV transmittances are expressed as a function of profile dependant predictors and regression coefficients, which are derived from a line by line model using a training set of atmospheric profiles. The coefficients provided with RTTOV are trained using the LBLRTM. However, due to the dependency of the RTTOV output on the line by line model used to train the regression coefficients, new coefficients have also been calculated using the RFM. This removes the differences between the underlying line by line model from comparisons.
Presented here are comparisons between the two models for a subset of atmospheric profiles, examining spectral regions in which there is good agreement and those that show large discrepancies. Further, the models are used to simulate real IASI measurements, using ERA-Interim profile data, to ascertain how well both models are able to realistically replicate the IASI spectrum.