Use of random numbers in computer simulation of magnetic resonance signals

The proton magnetic resonance spectra of water molecules diffusing over regular sites in natrolite are numerically simulated. Experimental data and the results of simulation are found to be in good agreement in a wide temperature interval including the transition range, where the magnetic resonance spectrum is a superposition of a wide unaveraged spectrum and a spectrum averaged by the motion of molecules. The simulation uses generators of pseudorandom-number Markovian sequences. For a binary Markovian process, it is shown, in particular, that the time within which a physical quantity remains constant is described by an exponentially distributed random number.