A ROBUST SATELLITE DATA ANALYSIS TECHNIQUE (RST) FOR SAHARAN DUST DETECTION AND MONITORING

A general satellite data analysis approach named RST (Robust Satellite Technique) has been successfully applied to monitor many natural and environmental risks. In this paper it is applied for the first time to Saharan dust detection by using optical satellite data collected in the visible (VIS) and thermal infrared (TIR) portion of the e.m. spectrum. The problem of identifying Saharan dust clouds, distinguishing them from small, low or thin, meteorological clouds is faced by combining spatial and spectral signatures in the visible and thermal infrared AVHRR remotely sensed radiances. Long term satellite records collected in similar (same time of the day, same month of the year) observational conditions have been used in order to characterize the desert background in terms of expected signal behaviour in the VIS (for spatial texture analyses) and TIR (for signal intensity analyses) spectral ranges in absence of dust clouds. Preliminary results achieved in the case of a dust-storms that hit North Africa on May 1997 suggest that the proposed technique, combining spatial and spectral signatures within the framework of the more general RST approach, could actually allow to distinguish between small, low or thin, meteorological clouds and dust clouds even over very variable surface backgrounds.