FIELD: physics.
SUBSTANCE: brightness temperature values are obtained over three radiometric channels having frequency of 19.35 GHz and 37.0 GHz with horizontal polarisation and 22.235 GHz with vertical polarisation. Cumulative humidity values are calculated using a relationship which takes into account brightness temperature values and coefficients a set-up neural network. Numerical values of coefficients of the set-up neural network which are included in the relationship for estimating cumulative humidity are obtained by mathematical modelling of the outgoing radiation of the Ocean-Atmosphere system and carrying out a numerical experiment using neural networks as an inverse solution operator with subsequent adjustment of the method on satellite and ground measurements which are superposed in space and time.
EFFECT: high accuracy of estimation, wider range of application conditions.
