RETRIEVAL OF ATMOSPHERIC PROFILES FROM RADIO OCCULTATION MEASUREMENTS USING AN ARTIFICIAL NEURAL NETWORK R. S. Cintra 1 , H.F de Campos Velho 1 and J. D. Simões da Silva 1 1 Laboratory for Computing and Applied Mathematics – LAC Brazilian National Institute for Space Research (INPE) Caixa Postal 515 – 12245-970 – São José dos Campos, SP-Brazil {rosangela.cintra,haroldo, demisio}@lac.inpe.br Abstract. A new method based on Artificial Neural Network (ANN) is applied to estimate high-resolution atmospheric profiles in the troposphere. The use of data satellites with the objective to study the atmosphere together GPS (Global Positioning System) data opens perspectives to improve the research on climate and the capacity on weather forecast. In this sense, many techniques were developed for retrieving atmospheric profiles (temperature, pressure and water vapor) using GPS radio occultation. ANN technique establishes a non-linear relation with the meteorological variables. The ANN retrieved profiles using meteorological variables from the CHAMP-ISCD (Challenging Mini-satellite Payload for Geoscientific Research and Application). It can be concluded that ANN is convenient and an accurate tool to get atmospheric profiles. 1. Introduction The atmosphere is made up of various gases that act as a protective shield for the Earth and allows life to exist. The atmosphere consists of five layers, held around the planet by the force of gravity. Moving upwards through the layers, the atmospheric pressure decreases rapidly with height and the air temperature also changes. The temperature variation is more complicated changes are used to divide the atmosphere into layers being the atmospheric air a mixture of gases constituted of dry added air of water steam (varying in space and time). The troposphere is the layer of the surface, the principal means of transport of mass (water, solid, pollutant particles, etc.) happens. The vapor transport in the atmosphere constitutes a very important component of the hydrological cycle, where a great quantity of water is transported in the form of vapor. The knowledge of the vertical and horizontal distribution of the water vapor in global scale is useful for applications of numerical weather forecast, climatic modeling and studies of climate global changes. Vapor pressure (vp) is the pressure of the vapor over a liquid (and some solids) in equilibrium; it is the fraction of the pressure of the environment due to the fraction of water vapor in the air. The superior limit for the quantity of water vapor in the air is a function of the temperature (T). By using remote sensor data of the atmosphere (from meteorological satellites) vertical soundings have been estimated the humidity and temperature vertical profiles. The use of low Earth orbit (LEO) satellites together Global Positioning System (GPS) data can also help on the study of the atmosphere and such procedures opens perspectives to improve research on climate and weather forecasting. In this sense, many techniques have been developed for retrieving atmospheric profiles (temperature, pressure and water vapor) using GPS radio occultation. 2. Radio-Occultation data Radio-occultation (RO) is the phenomenon for which a celestial body can be seen due to the occlusion of another celestial body. The atmospheric soundings of the temperature and of the water vapor are retrieved from the measures of RO from the GPS. The Figure 1 shows the behavior of the