Incorporation of geological constraints into geochemical anomaly identification using BME-GWR: A case study from Inner Mongolia of China

Abstract

Delineation of geochemical anomalies is a fundamental task in mineral and environmental investigation. Mineralization area of interest is mainly controlled by underlying geological environment. These geological constraints facilitate to identify mineralization anomalies and remove pseudo-anomalies. Regional geochemical data is one of most commonly used in deriving mineralization information, which is also considered to be the primary data for direct geochemical anomaly identification. The selected study area, Jining in central Inner Mongolia of China, is mostly covered with Quaternary sediments, whereas is located in the important molybdenum polymetallic mineralized belt. The overburden shields the geochemical signatures from deep, resulting in missing some important information. The NE-trending faults, intersection of faults, Jining formation and mid-acid granites is mainly ore-forming factors in this study area, which are used as the geological constraints for identifying molybdenum polymetallic mineralization. Geographical weighted regression (GWR) generates the mean and variance of probability soft data based on these geological constraints. Bayesian maximum entropy (BME) integrates the constructed soft data with geochemical data for obtaining practical mineralization anomalies. These results shown that the combination of BME and GWR can identify local anomalies and weak anomalies compared with the results obtained from geochemical data alone. Specially, some anomalies of different periods of deposits have been highlighted in the results of BME-GWR model. The discovered deposits correspond tothe identified mineralization anomalies well. The BME-GWR model provides a potentially powerful tool for geochemical anomaly identification under consideration of geological constraints.