M.I. Shimelevich, E.A. Obornev

Russia State Geo Рrospecting University, e-mail: [email protected]

The neuronet (NN) method of estimation of time changes (monitoring) of dynamic geoelectrical section parameters on the base of incomplete magnetotelluric data is developed. Incomplete data are the results of measurements of the electromagnetic fields, carried out on the base of rarefied spatial observation networks, in particular, localized in fixed domains of the ground surface. Constructed NN inverse operators of special type allow to make fast inversion of data and to observe the changes of fixed parameters of media in a mode of real time. Work of algorithm is illustrated on synthetic 2D magnetotellurics data.