Several methods are used to retrieve cloud top pressure from remote sensing data. Methods based on IR observations sense the temperature of the cloud top and yield the pressure once the temperature profile of the atmosphere is known. Methods based on the O2-A absorption band near 477 nm, essentially measure the amount of oxygen above, and partly inside, the cloud. Relatively new methods are those based on Raman scattering, i.e. the filling in of Fraunhofer lines in the earth-shine spectra due to N2 and O2 above the cloud, and methods based on absorption by the O2-O2 collision complex. The last methods is currently being developed and tested, so that it may be used to obtain cloud cover and cloud top pressure from spectra recorded by the Ozone Monitoring Instrument (OMI), which will be launched in 2003 together with the other instrument of EOS-AURA. As OMI has a spectral range of 270-500 nm, cloud top pressures can not be derived from IR and O2 A band observations, and at least one of the relatively new methods has to be used.

In this presentation we consider the algorithm based on O2-O2 absorption. Results of sensitivity studies using synthetic spectra calculated with the doubling/adding method will be shown. For comparison purposes, the algorithm has been applied to spectra obtained with the Global Ozone Monitoring Experiment (GOME) which also observes the O2-A band. This allows us to compare the results of our algorithm with that based on the O2-A absorption band. A preliminary conclusion is that the weak absorption feature of the O2-O2 collision complex at 477 nm is well suited to derive pressures that characterize the cloud. However, the method becomes less accurate for high-altitude clouds, because absorption by the O2-O2 collision complex is then very weak.