Application of the correlation theory of thermal regime and thermal radio emission for atmosphere

The results obtained by application of the earlier developed stochastic theory of temperature distribution and thermal radio emission of medium (half-space) to atmosphere are presented. It was obtained that the exponential autocovariance function of the surface temperature is in good agreement with that determined from meteorological data (such covariance functions are inherent in random processes generated by the Poisson process). The theory offers a good qualitative explanation of the frequency and height dependencies of statistical parameters, and for some of these parameters there is a good quantitative agreement. The theoretically predicted effect of the time shift of the correlation functions maxima was discovered in the data. The theory was generalized for the spatial inhomogeneities of atmosphere, and the vertical correlation length of random atmosphere inhomogeneities was obtained as the geometric mean from the diffusion length and the horizontal correlation length of surface temperature.