Cloud detection scheme for numerical weather prediction assimilation of IASI data

Abstract

Measured cloud spectra have been used to develop an initial cloud detection scheme for the assimilation of Infrared Atmospheric Sounding Interferometer (IASI) data. Using realistic data before the instrument is in orbit will help ensure that IASI data can be used at numerical weather prediction (NWP) centers soon after launch. The measured data have been taken by the Airborne Research Interferometer Evaluation System (ARIES) - specifically designed to gather data representative of IASI from an aircraft. The development of the cloud detection scheme and its final form are discussed. The scheme was designed to require no additional background information. Representative spectra are used to create a set of empirical orthogonal functions (EOF). EOF sets were based on 78 diverse modeled clear air spectra, supplemented by selected measured spectra. IASI data could be compressed using such EOF. Tests were carried out on various cloud detection procedures using separate measured spectra. Video data gave independent verification of cloudy and cloud free views. The results were encouraging; clouds were detected in the measured test data with the prospect of extending the results to more general data.