where deta is the difference between the dataset of interest and the reference data, n is the number of samples and ME the mean error.
the accuracy of a dem is usually described by only one global accuracy parameter for the entire dataset. kraus et al.(2006) state insufficient because error might correlate significantly with difference type of surface characteristic (e.g.flat,hilly or mountainous areas). therefore, the authors claim that the spatial distribution of the error has to be taken into account when describing geospacial datasets (hohle and hohle,2009)