I have done autocorrelation for experimental data ... That was easy enough since I had time series data.
But what about this spatial correlation? Wouldn't I need full spatial solution for a very large number of timesteps? Did you do it using a runTime function object?