Remote Sensing of Water Cloud Parameters Using Neural Networks

Abstract

Abstract In this work a method for determining the micro- and macrophysical properties of oceanic stratocumulus clouds is presented. It is based on the inversion of a radiative transfer model that computes the albedos and brightness temperatures in the NOAA Advanced Very High Resolution Radiometer (AVHRR) channels. This inversion is performed using artificial neural networks (ANNs), which are trained and optimized by genetic algorithms to fit theoretical computations. A detailed study of the ANN parameters and training algorithms demonstrates the convenience of using the “backpropagation with momentum” method. The proposed retrieval method is applied to daytime and nighttime imagery and was validated using ground data collected in Tenerife (Canary Islands), obtaining a good agreement.