Evaluation of soil fertility is a key factor determining the level of fertilizer need. Application of chemical fertilizers may unbalance the soil nutrient, damage the soil environment with no effect on yield quality and quantity, if the soil fertility level is not considered. In order to optimize the use of soil resources, understanding the spatial variability of soil properties is necessary. The study of spatial variability of soil properties provides information on the spatial distribution of soil properties as a main tool for researchers, policy makers and managers dealing with different soil aspects. In this work the spatial variability of some soil chemical properties was studied to map the soil fertility in 50,000 hectares of Lenjanat region, Isfahan province. Using a systematic grid sampling with 500 m intervals, 1180 compound surface samples (0-25cm) were taken by the Department of soil and water in Isfahan Agricultural and Natural Resources Research Center and analyzed for nitrogen, phosphorus, potassium and organic carbon. The spatial distribution of the measured properties, first with the Kriging interpolation and inverse distance weighting (IDW) methods and then by combining fuzzy logic Kriging were analyzed. The spatial structure of variables was analyzed by the means of the variogram. Omnidirectional and directional variograms were calculated and plotted for all variables. The results showed a strong spatial structure for the variables exceptional of P. Moreover, the spherical model showed the best fit for all variables. The anisotropy of variables were studied using directional variogram in three main directions (0, 45, 90 degrees) with a tolerance angle of 22.5?. The studied variables showed a geometric anisotropy. To check the ability of the estimation techniques, some standard and statistical parameters such as Mean Estimation Error (MEE), Mean Square Estimation Error (MSEE), Root Mean Square Estimation Error (RSME), and Pearson correlation coefficients were used. The results showed a high accuracy of the estimated values for N, K, and organic carbon. But for P the Pearson correlation coefficient for both methods of ordinary kriging interpolation and inverse distance weighting showed a low accuracy of estimation. Since the spatial distribution of the individual data does not show the fertility status of the soil, the fuzzy logic was applied to combine the se factors. With the mean of fuzzy logic, maps of the individual nutrients in soil were combined and a soil fertility map was developed. The map described four Keywords: Geostatistics, Lenjanat, Fuzzy Logic, Mapping LandFertility
