Autocorrelation is the propensity of data values to be similar to surrounding data values. In a raster image, this describes the degree to which values in any cell will be similar to the cellsvsurrounding it. A common measurebof autocorrelation is Morans I, with a range from -1 yo +1. A value close to +1 indicates a smooth surface, with each cell containing a value very similar to the cell next to it. A value close to -1 indicates a very rough or fractured surface, with each cell quite from its neighbors.different