Agricultural fields close to Rabia district, northwest of Nineveh Governorate / Iraq were selected for the study, as the study area is located between the points (36°31'51.34'' and 36°43'40.343'') north and the points (24°16'14.475'' and 42° 34'50.99'') east, and the study area was approximately 52.5 hectares (210) km2. Predictive models were built resulting from multiple linear and nonlinear regression relationships between remote sensing data and ready concentration values of the three necessary components (N, P and K) measured in the laboratory, in order to apply the predictive model to space data for previous years and assess the state of deterioration in soil fertility for two decades of time. The results of the study showed the possibility of applying predictive models to space data for previous years to give results with a high spatial accuracy of (0.9713, 0.9492 and 0.9670) for nitrogen, phosphorous and potassium, respectively. Spatial maps could be drawn for each of the three studied years (2002, 2012 and 2022) and for each of the three studied nutrients. The amount of deterioration that occurred and a decrease in the concentration of nutrients over time was observed as a result of bad management and poor exploitation of the earth's resources

nitrogen, phosphorous, potassium, soil fertility, remote sensing, predictive models.