The problem of synthesizing an algorithm for the identification of optometeorological situations describable by parametric Gaussian noise curves is considered within the framework of Bayesian approach. The proposed procedure is capable of adaptation; to real atmospheric situations.