1: \begin{abstract}

2: We present here an explicit form of the random spectral measure element, what allows us to express a stationary random field as a stochastic integral explicitly depending on its power spectrum and a spectral tensor if the field is a vector one. It has been shown here that convergence mechanism of such integral is significantly different from the one of the Fourier transform and that the traditional formalism is a partial limiting case of the one presented here. The fact that there is an explicit expression of a random field makes calculation of higher order statistics of it much more straightforward (see for example \citealp{Ch20}). For a vector field such expression contains a projection of an isotropically distributed random vector by a spectral tensor, what makes geometrical interpretation of harmonics behavior possible, simplifying its analysis (see \S\ref{sect:vf}). This spectral representation also makes straightforward numerical generation of a random field, what is extensively used by \cite{Ch20}. We also present here some practical applications of this formalism.

3: \end{abstract}

4: