In this case study the geostatistical analysis of digital hydrogeological map accuracy was carried out. Maps were modeled in isolines by means of different interpolators. The occurrence of groundwater levels was mapped on the representative area in limits of the Republic of Moldova. Using the software Surfer 11 eleven models of groundwater levels were generated by the following interpolators: Kriging, Radial Basis Function, Inverse Distance to a Power, Modified Shepard's, Minimum Curvature, Polynomial Regression, Triangulation with Linear Interpolation, Nearest Neighbor, Natural Neighbor, Moving Average and Local Polynomial. The comparability between maps was estimated applying methodology of the statistical residuals as the subtraction between Natural model –Artificial model and the absolute average value of the statistical residuals. Modified Shepard's method is defined as the most accurate interpolator. Nevertheless the correlation analysis shows that this interpolation method is very close with Kriging, Local Polynomial, Minimum Curvature, Natural Neighbor and Triangulation with Linear Interpolation (r> 0.70).

În studiile prezente a fost analizată precizia hărţilor hidrogeologice digitale, care au fost produse folosind metodele diferitor interpolatori în izolinii. Pentru un teritoriu reprezentativ, în cadrul R.Moldova, s-a cartat nivelul apelor freatice în condiţii naturale. Utilizînd softul Surfer 11 au fost întocmite 11 modele a nivelului freaticului prin intermediul interpolatorilor Kriging, Radial Basis Function, Inverse Distance to a Power, Modified Shepard's, Minimum Curvature, Polynomial Regression, Triangulation with Linear Interpolation, Nearest Neighbor, Natural Neighbor, Moving Average şi Local Polynomial. Diferenţa între hărţile obţinute a fost estimată folosind metodologia reziduurilor statistice ca diferenţa dintre Model natural –Model artificial şi a valorii reziduurilor absolute medii. Cel mai exact interpolator a fost considerat Modified Shepard's Method. Tot odată în rezultatul analizei corelaţionale această metodă este foarte apropriată de interpolatorii Local polynomial, Kriging, Minimum curvature, Natural neighbor şi Triangulation with linear interpolation (r >0.70).