Cloud detection over sea surface by use of autocorrelation functions of upwelling infrared spectra in the 800–900-cm^-1 window region

Abstract

A new algorithm for cloud detection over sea surface has been developed that makes use of autocorrelation and cross correlation between a real spectrum and either a synthetic or a laboratory spectrum. The scheme is intended for high-spectral-resolution satelliteborne infrared sensors that will measure the Earth's entire emission spectrum rather than the upwelling radiance in a few channels. A new index is defined with which one can determine quantitatively the degree of homogeneity of two spectra in the 800-900-cm(-1) (11.11-12.5 mum) window region. The index makes use of only the observed spectrum along with a reference synthetic spectrum and, therefore, may form the basis for an operational stand-alone cloud-detection algorithm for next-generation high-spectral-resolution infrared sensors. Application both to synthetic spectra obtained in simulation and to sea-surface real spectra recorded through the interferometric monitor for greenhouse gases is considered.