Classification and analysis cloud types using multispectral data of the radiometer AVHRR of the satellite NOAA-14

The clouds play an important role in the radiative balance in the top of atmosphere. They reduce the absorption of the solar radiation due to the increase of the reflectivity, and produce a reduction of the infrared radiation for being opaque to the terrestrial radiation. Because of this reason the classification and determination of the amount of cloud have been one of the most important tasks in Remote Sensing. The different cloud types have different radiative properties (emissivity, transmissivity and reflectivity) which change in agreement with the density, thickness, height and phase of the particles. The AVHRR (Advanced Very High Resolution Radiometer) radiometer of the polar orbit satellites of the series NOAA-n has spectral channels in the pseudo atmospherics windows . With the advance of the multispectral images and the technological, it became possible the utilization of objectives technique to extract desirables parameters of the sensors of satellites. Algorithms using spectrals bands have been developed in order to obtain of several parameters, mainly to provide statistics information of the clouds for the use in climatological analysis and delimitation of areas without clouds. The purpose of this work is to develop a method of classification and determination of coverage of clouds using a multispectral approach from data of AVHRR. However, to maintain the quality of the operational product is necessary a frequent validate. To accomplish this, is necessary a development of tools of image analysis and direct visualization (photo interpretation) along with comparisons to surface synoptical observations. The digital color image produced by algorithm can be compared to the satellite original image. At the same time, the amount of clouds layer, the coverage of each cloud layer, type and others results of the algorithms can be compared to other separater algorithms of clouds that are not able to produce digital images. As soon as validation is done with surface observations and others meteorological observations, this method can turn into operational and available to scientific community.