Ordinary kriging assumes that the mean of the variable is constant across the study area. This method also assumes that the spatial correlation between data points can be modeled by the variogram. It is important that the data exhibit spatial autocorrelation, meaning that closer points are more similar than those further apart.
