Soundings
6. Giuseppe Grieco, Guido Masiello and Carmine Serio
|
|
As a part of the tools, which the applied spectroscopy Group at DIFA (Department of Environmental Engineering and Physics) has developed mostly to assist the development and commissioning phases of IASI, an end-to-end standalone retrieval package for nadir sounding instruments has been implemented and released. It is called Φ-IASI, which reflects the original purpose for the code, but may be used with many instruments (it is currently set up for IMG, NAST-I, AIRS and IASI, and future sensors for geostationary platforms, such as MTG-IRS, but others may be easily added).
The package Φ -IASI is intended to provide a kit of models to address research issues on inversion methodology (including Tikhonov and/or Rodgers regularization, Levenberg-Marquardt least-square minimization) and radiative transfer (including generation of analytical derivative matrices, impact of new spectroscopy). The package has been the subject of various scientific papers and, in addition, it has been extensively validated using aircraft and satellite high spectral resolution infrared observations recorded with Fourier transform spectrometers
The Φ-IASI package includes a forward/inverse radiative transfer model, a cloud detection scheme and a linear statistical retrieval approach for the computation of the first-guess state of the atmosphere, which is used for the initialization of the nonlinear physical inversion.
Cloud detection
The cloud-detection scheme we have implemented in Φ-IASI is a stand-alone scheme that uses only the observed spectral radiances. The scheme is based on a
series of various threshold-tests and is trained by using a suitable dataset of clear-sky spectral radiances.
Statistical retrieval
The Φ-IASI methodology includes a statistical retrieval scheme which is normally used to provide the first-guess state of the atmosphere to initialize the
nonlinear physical inverse scheme. The principal component (PC) or empirical orthogonal function (EOF) transform is used to decompose the data (spectral radiance)
and the parameters (temperature, water vapour and ozone profiles) into two new orthogonal bases, one for the data and one for the parameters. A linear regression
relation is then established between data and parameter PC scores. An important characteristics of our regression scheme is that it is fully analytical, that is
the regression coefficients are analytical and they have been derived for the case of a generic signal-noise model, the methodology can therefore be used to
design and implement retrieval algorithms according to the user needs.
The forward model σ-IASI.
This consists of a monochromatic radiative transfer model which has been designed for the fast computation of spectral radiance and its derivatives (Jacobian)
with respect to a given set of geophysical parameters. The forward module computes monochromatic radiances from look-up tables of monochromatic-layer optical
depth generated using the line-by-line model LBLRTM.
The inverse model δ-IASI
The inversion package incorporates its own forward module (i.e. σ-IASI), therefore it can be considered as an end-to-end procedure, which, given the
observed spectral radiance as input, returns the geophysical parameters without any further intervention by the user. The inverse module, δ-IASI
implements a nonlinear inversion procedure and, therefore, it has to be properly initialized. The initialization is provided by the EOF statistical retrieval
approach.
Further details and appropriate references can be found at the Φ -IASI web page below
http://www.difa.unibas.it/jFM/dlf/Laboratori/ApplSpec/as/phi.html
The figure below shows an example of retrieval products for temperature, water vapour and ozone, obtained from IASI radiances (six spectra) recorded
on occasion of the JAIVEx experiment (http://badc.nerc.ac.uk/data/jaivex/)
Example of phi-IASI retrieval. The example refers to six IASI spectra recorded on 29 April 2007, during the JAIVEx experiment. For ozone the comparison
is with the ECMWF analysis, dropsonde for the case of temperature and water vapour (details in Masiello et al, ACP, 2009,
www.atmos-chem-phys.net/9/8771/2009/