This study is an attempt to understand the application of spectral image processing methods to ASTER data for mapping host rock associated with porphyry copper-molybdenum mineralization. The application of remote sensing in identifying variations in surface mineralogy, structural elements, and lithologic contacts can help in identifying such relations. Signatures collected from the Advanced Spaceborne Thermal Emission Reflection Radiometer (ASTER) data is used to map the host rock of Zafarqand area. The study area is located in the central segment of the Urumieh–Dokhtar volcanic belt of Iran. The Urumieh-Dokhtar is a potential zone for exploration of new porphyry copper deposits. Band ratio, band math, false colour composites and principal component analysis techniques are used to refine the different lithologic units in the area. These methods showed the discrimination of acidic igneous rock from intermediate igneous rock and the boundary between igneous rocks using the shortwave infrared bands (SWIR) of ASTER at regional scale. Results have proven to be effective and in accordance with the results of field observations, ore microscopy examination and with reference geological map. In conclusion, the image processing methods used can provide cost-effective information to discover possible locations of porphyry copper and epithermal gold mineralization prior to detailed and costly ground investigations. The extraction of spectral information from ASTER data can produce comprehensive and accurate information for copper and gold resource investigations around the world, including those yet to be discovered.
Generative adversarial networks (GAN) and various deep autoencoders have been frequently executed to recognize multi-element geochemical anomalies linked to different ore resources in recent decade. Efficient recognition of multi-element geochemical anomaly patterns is a significant issue in mineral exploration targeting. Traditional procedures have not sufficient capability to perform efficient pattern recognition. While, deep learning algorithms as influential subset of machine learning algorithms can present magnificent conclusions in classification and pattern recognition. Because those have robust ability in extracting high-level features of complex inputs. Although, many deep learning algorithms were used to recognize geochemical anomalies but the GANs have demonstrated specific dignity in recognizing multi-element geochemical anomaly patterns. But, these frameworks should be constrained to learn geological knowledge and yield reasonable potential maps. In this regard, a novel geologically-constrained GANomaly was trained with frequency domain training data to recognize multi-element geochemical anomalies. Application of the geologically-constrained GANomaly network with considering mineral system parameters of the Au–Cu mineralization in the Feyzabad district, NE Iran was eventuated to suitable results. The success-rate curves demonstrated that produced map of frequency domain geochemical data has traced 86.68% Au–Cu occurrences via 30% corresponded area while produced map of spatial domain geochemical data has traced 80.13% Au–Cu occurrences via 30% corresponded area.
Reliable recognition of geochemical anomalies linked to ore deposits is one of the most significant challenges in mineral exploration. Several advanced machine learning (AML) algorithms have recently been applied to recognize multi-element geochemical anomalies. Performance of the AML algorithms are extremely dependent to values of their hyperparameters. Because, conclusions of their application can significantly be differed tuning hyperparameters. Tuning hyperparameters through trial-and-error way is a labor-intensive and time-consuming procedure which is not mostly eventuated to reliable results. In this regard, applying an AML model decreases training time and assists to achieve optimized values of hyperparameters yielding reasonable potential maps. Hence, execution of an AML model mitigates the biasness problem and uncertainties with recognition of multi-element geochemical anomalies. In this study, Harris hawks optimization (HHO) algorithm was employed to optimize known hyperparameters of the random forest (RF) method for detecting multi-element geochemical anomalies related to mineralization occurrences in the Feyzabad district of the Razavi Khorasan province, NE Iran. This research demonstrates that Harris hawks optimized random forest (HHORF) model is a vigorous procedure to identify multi-element geochemical anomalies. Because, the HHORF model has recognized 86.53% mineralization occurrences through 30% corresponding area while the RF method has catched 80.14% mineralization occurrences up via same corresponding area.