Geostatistical analysis and interpretation of Ilesha aeromagnetic data south–western, Nigeria

Abstract

The uses of variogam and kriging as a tool in geostatistical analysis have gained greater prominence recently in the diverse scientific field, especially for mineral exploration purpose. Ilesha, the study area, has been identified as the one the region with abundance gold deposits in Nigeria. Different methods have been used in the past for the analysis and interpretation of aeromagnetic data in the gold deposit area with less attention to the geostatistical approach. The objectives of this work are to (i) fit the aeromagnetic data into the variogram model to estimate the magnetic spatial structural dependency on the geological composition (ii) delineate the spatial magnetic anomaly associated with the lithological units using kriging interpolation techniques (iii) deduce the zone associated with strong and weak gold mineralization (iv) evaluate the kriging techniques for cross validation. The major tool used in this work is the geological map of Ilesha which was partitioned into nine (9) lithological H-units in conjunction with an aeromagnetic sheet obtained from the Nigeria Geophysical Survey Agency, Abuja. In this study, three variogram models, the spherical (S), exponential (E) and Gaussian (G) models, were used. Three kriging interpolation techniques, ordinary kriging (OK), co-kriging (CK) and radial basis function (RBF) were employed. Nugget Sill Ratio (NSR) was deduced to estimate the autocorrelation level of the variogram models while cross validation was carried out on the kriging techniques using mean square error (MSE) and root mean square error (RMSE) for predictive performance evaluation. The result obtained accounted for the variogram model in the order of S < E < G for six (6) lithological H-units. Two units (H3 and H4) were found in the order of E < S < G while one unit, H5 is in the order of S < G < E. The NSR result revealed two distinct levels, namely; a strong and moderate level. Six H- units fall under the strong autocorrelation dependency in the range of 6.78–20.79%, while three H-units have moderate autocorrelation dependency within the range of 26.07–58.91%. The kriging results accounted for three distinct magnetic anomalies; low, moderate, and intense, across the nine lithological H- units. The gold strong mineralization zone are attributed to hydrothermal alteration in the region with low to moderate magnetic field intensity in the range  < − 100 nT to 50 nT. The interpolation performance evaluation revealed CK to have the lowest MSE and RMSE value when compared to OK and RBF. The three kriging techniques adopted are good linear predictive estimator but CK gives a better predictive accuracy and have less perturbation. In this study, the application of the geostatistical methods (variogram and kriging) to the analysis of Ilesha aeromagnetic data has led to the conclusion that these techniques are effective mathematical tools for delineating the structural and spatial dependency of magnetic anomalies which have a great attribute in mineral exploration.