Exploiting quartz spectral signature for the detection of cloud-affected satellite infrared observations over African desert areas

It is shown that IMG (interferometric monitoring of greenhouse gases) spectra recorded over African and Arabian deserts clearly contain the fingerprint of quartz-rich soils. We illustrate how this spectral signature can be exploited to devise a suitable cloud-detection scheme to identify which infrared observations are affected by clouds. As a by-product, the scheme also allows one to identify the most likely underlying emitting surface type and provides a suitable first guess for the surface emissivity to be used, e.g., for the retrieval of geophysical parameters from high-spectral-resolution infrared radiance from space. The analysis has focused on African deserts because of their intrinsic relevance to numerical weather prediction and Earth's climate. Desert areas, like oceans, are poorly covered by the world meteorological radiosonde network and therefore are geographical regions for which the global coverage capability of satellites soundings is expected to provide better initializations for numerical weather prediction than are now available. Application of the cloud-detection scheme to IMG spectra has been considered, which demonstrates the good performance of the method.