A new study published in Discover Geoscience by Daniela Kuranaka, César Augusto Moreira and colleagues at São Paulo State University, the Federal University of Pampa and the University of the Republic has now taken up that challenge with an integrated toolkit that blends old-fashioned field geology with drone photogrammetry, ground magnetometry and gamma-ray spectrometry. Their results point to a specific culprit: a shear zone slicing through the Cerro da Cria Granite, where hydrothermal fluids appear to have altered the rock, destroyed its magnetite and left behind a potassium-rich fingerprint that matches the signature of structurally controlled orogenic gold systems worldwide.

The geological stage for this detective story is the South Brazilian Shield, a roughly 50,000-square-kilometre expanse of Precambrian crust that records more than two billion years of Earth history. Its present architecture was largely forged during the Brasiliano Orogeny, a protracted episode of mountain building between about 900 and 535 million years ago driven by the collision between the Kalahari and Rio de La Plata cratons. Within the Dom Feliciano Belt that resulted from this collision, the São Gabriel Terrane preserves the remains of Neoproterozoic intra-oceanic magmatic arcs, with zircon ages clustered between roughly 880 and 720 million years. Later, as the orogen collapsed, extensional tectonics opened the Camaquã Basin and unleashed felsic volcanism of the Acampamento Velho Formation, dated between 580 and 570 million years ago, while pre-existing structures were reactivated deep in the crust.

Three main rock units crop out at Cerro da Cria, and each tells part of the story. The Vacacaí Metamorphic Sequence comprises orange phyllites and metavolcanic rocks metamorphosed under low- to medium-grade conditions, relics of the ancient arc and back-arc environments. The Acampamento Velho Formation contributes brownish to reddish rhyolites with small quartz phenocrysts, products of the late post-collisional volcanism. Cutting discordantly across both is the Cerro da Cria Granite, a reddish, equigranular body of quartz, feldspar and biotite emplaced during post-collisional magmatism. Crucially, the whole terrane is threaded by regional northeast-southwest shear zones developed under dextral transpression, including the Palma–Vila Nova do Sul Shear Zone, which governed the elongation of the metavolcanic sequences and the emplacement of the granite itself. In such settings, these zones of weakness are exactly where mineralizing fluids tend to travel.

To test whether the stream-sediment gold could be traced back to a primary source, the team designed a campaign that covered approximately 70 hectares of the basin in just five days of fieldwork. A DJI Mavic 3T drone flew two missions at 80 metres altitude and 7.5 metres per second, capturing imagery that was processed in Agisoft Metashape into a high-resolution orthophoto and digital elevation model referenced to the SIRGAS 2000 geodetic system. On the ground, the researchers established a georeferenced grid with 25-metre spacing using differential GPS accurate to roughly half a metre, yielding 837 measurement stations. At each station they simultaneously recorded gamma-ray spectrometry with an RS-332 instrument equipped with a bismuth germanate detector, and total-field magnetic measurements with a GSM-19 proton precession magnetometer. A base station logged the Earth’s magnetic field every two minutes so that diurnal variations could be stripped from the rover data, isolating the signal of the subsurface.

The gamma-ray spectrometry exploits a simple physical principle: potassium-40, and the decay chains of uranium-238 and thorium-232, emit gamma photons of characteristic energies that scintillation crystals convert into measurable electrical pulses. The resulting maps of the three radioelements proved remarkably diagnostic. Potassium concentrations ranged from 0.00 to 3.80 percent, with the strongest anomalies in the central-western sector reflecting the granitic body itself, and a secondary enrichment zone of 0.97 to 1.51 percent in the northern part of the area interpreted as the influence of the shear zone. Thorium, spanning 1.17 to 10.90 eppm, broadly mirrored the potassium pattern, while uranium showed a more irregular, patchy distribution consistent with remobilization along microfractures and weathering profiles. The potassium-poor mafic metavolcanic units, by contrast, stayed below one percent, confirming that the radiometric contrast cleanly discriminates the main lithologies.

The magnetic data added a second, complementary layer of evidence. Total magnetic intensity across the survey area varied from 22,155 to 22,686 nanotesla, a 531-nanotesla amplitude, with the highest values in the east and southeast over the granite and lower values toward the northwest where the metasedimentary rocks of the Vacacaí Sequence begin. Mathematical processing, including reduction to the IGRF reference field and calculation of the Analytic Signal Amplitude and vertical derivative, sharpened the picture. The ASA, which ranges from 0.04 to 40.68 nanotesla per metre, highlighted strong magnetic contrasts in the northwestern and southeastern portions of the granite, while the vertical derivative, spanning -2.41 to 2.13 nanotesla per metre, revealed short-wavelength lineaments predominantly oriented east-west. These lineaments segment the granite and correspond to brittle-ductile fracture zones, some of which even express themselves in the topography captured by the drone.

When the datasets were overlaid, the story snapped into focus. In the northeastern sector of the basin, potassium enrichment coincides with structural features highlighted by the analytic signal, and the combination of radiometric anomalies with comparatively reduced magnetization is the classic signature of potassic hydrothermal alteration: fluids percolating along fractures and shear zones chemically transform and partially destroy magnetite, lowering the rock’s magnetic response while flooding it with potassium. The team interprets the increased permeability along these structures as the engine that drove potassium-rich hydrothermal circulation, accompanied by the concentration of oxides and sulfides. Field structural measurements reinforced the interpretation: 84 foliation measurements show a dominant northeast-southwest fabric produced by ductile deformation during the Brasiliano cycle, while 37 joint measurements cluster into four sets oriented northwest-southeast, east-west, north-south and northeast-southwest, recording late brittle reactivation. At one outcrop, a fault plane in graphite-talc schist with strongly stretched minerals defines the Cerro da Cria Fault, a sinistral strike-slip shear zone trending east-northeast to west-southwest that the authors identify as a principal control on fluid circulation and gold distribution.

This geophysical and structural fingerprint aligns closely with the metallogenetic model of orogenic gold systems, in which gold precipitates from hydrothermal fluids channeled through crustal shear zones during transpressional deformation. In the Southern Brazilian Shield, such mineralization is typically hosted in quartz veins and alteration zones within metavolcanic, metasedimentary and granitic rocks, commonly associated with pyrite and arsenopyrite. The spatial correspondence between the radiometric anomalies, the magnetic response and the mapped shear zone led the researchers to conclude that the Cerro da Cria shear zone, with potential regional continuity along the Palma shear system, is the most likely primary source of the gold now scattered through the basin’s stream sediments, having been eroded from higher ground and transported downslope into the drainage network. Notably, the data offered no direct evidence that the granite itself is the gold source; rather, it served as the structural host that the fluids exploited.

Beyond the specific discovery, the study makes a persuasive methodological case. Semi-regional ground magnetometry and gamma-ray spectrometry, combined with drone-based photogrammetry and structural mapping, delivered detailed coverage of 70 hectares in five days, demonstrating a rapid, versatile and high-resolution alternative to widely spaced airborne surveys that can miss small exploration targets. The approach converts a regional geochemical anomaly into a concrete, prioritized drill-ready target, and the authors recommend trenching orthogonal to the identified structures to obtain larger samples for grade estimation, with induced polarization and electrical resistivity tomography suggested for concealed targets. As global gold consumption reached a record 4,974 tonnes in 2024 amid sustained central bank purchases, techniques that can cheaply and quickly narrow the search in structurally complex shields are likely to become essential tools, and the quiet hills of Cerro da Cria may prove to be an early showcase of how modern integrated exploration finds what centuries of prospecting overlooked.

Subject of Research: Integrated geophysical and structural modeling to locate the primary source of a gold geochemical anomaly in the Cerro da Cria drainage basin, southern Brazil

Article Title: Geophysical modeling in a drainage basin with a gold geochemical anomaly, Cerro da Cria Region, Vila Nova do Sul, Rio Grande do Sul (Brazil)

Article References: Kuranaka, D., Moreira, C. A., Barros, M. F. S., Carlos, D. U., de Oliveira, J. P. P., Ilha, L. M., Kumaira, S., & Masquelin, H. (2026). Geophysical modeling in a drainage basin with a gold geochemical anomaly, Cerro da Cria Region, Vila Nova do Sul, Rio Grande do Sul (Brazil). Discover Geoscience, 4(1), Article 291. https://doi.org/10.1007/s44288-026-00669-1

Image Credits: AI Generated

DOI: 10.1007/s44288-026-00669-1

Keywords: gold exploration, geophysics, gamma-ray spectrometry, magnetometry, shear zone, orogenic gold, Brazil, drone photogrammetry, hydrothermal alteration, geochemical anomaly, structural geology, mineral exploration

The work matters because the Birimian terrane of West Africa is one of the world’s great gold provinces, hosting world-class deposits such as Obuasi on the Ashanti Belt, yet many parts of it remain under-explored in a modern, quantitative sense. Rather than letting an algorithm sift raw data blindly, the researchers anchored their modelling in a mineral-systems approach. This framework, widely used in contemporary exploration targeting, asks explorers to translate what is known about how a deposit type forms—its sources of fluid and metal, the pathways that move those fluids, the structural traps where gold precipitates, and the alteration haloes left behind—into explicit, testable expectations. At Beposo, the documented structural and hydrothermal controls on gold mineralisation were converted into a set of geoscientific predictor layers against which the model outputs could be judged.

Those predictors were derived exclusively from airborne geophysical data: eight layers extracted from magnetic and radiometric surveys supplied by the Ghana Geological Survey Authority. Airborne magnetics reveals patterns of magnetisation in the crust, exposing faults, shear zones, lithological boundaries and intrusive bodies even beneath tropical soils and vegetation. Radiometric data, which measure gamma-ray emissions from potassium, thorium and uranium in the near-surface, serve as a proxy for hydrothermal alteration—because fluids that deposit gold typically also redistribute potassium and other elements in the wall rocks around a vein system. Crucially, the team did not feed geochemistry into the models as predictors; geochemical data were used solely to label the 67 known gold occurrences as positive training examples, a design choice that keeps the model honest and makes the resulting maps genuinely predictive rather than merely descriptive.

Two classifiers were trained and compared. The decision tree, a long-established technique dating back to the 1980s, makes predictions by recursively splitting the data according to simple rules, producing a transparent, tree-like logic that geologists can inspect. The random forest, introduced by Leo Breiman in 2001, builds hundreds of decision trees on random subsets of the data and predictors, then averages their votes. This ensemble strategy typically suppresses the overfitting to which single trees are prone and yields more stable probability estimates. For both models, the hyperparameters—the internal settings that govern how aggressively a model learns—were optimised using a 10-fold GridSearchCV procedure within a spatially independent training partition, a detail that matters because spatial autocorrelation in geological data can otherwise inflate apparent performance when neighbouring points leak between training and validation sets.

Evaluation was carried out on an identical held-out test set using a battery of complementary metrics: receiver operating characteristic and precision–recall curves, standard classification scores, and the Brier score, which penalises overconfident wrong probabilities. The random forest model achieved an area under the ROC curve of 0.93, an F1-score of 0.85 and a Brier score of 0.134. The decision tree trailed with an AUC of 0.85, an F1-score of 0.81 and a Brier score of 0.152. An AUC of 0.93 means that if one known gold occurrence and one randomly chosen non-occurrence location were drawn from the map, the model would rank the true occurrence higher 93 percent of the time—a strong result for prospectivity mapping, where the geological signal is notoriously subtle and the number of known deposits is small.

The authors are careful, however, not to overclaim. Because the held-out test set is small, the bootstrap confidence intervals of the two models overlap, meaning the random forest cannot be declared statistically superior to the decision tree. It is reported instead as retaining the higher central estimate on every metric. This kind of statistical candour is refreshing in a field where machine learning headlines often outrun the evidence, and it reflects a broader maturation of prospectivity mapping as a discipline: the question is no longer only whether an algorithm can fit the training data, but whether its probabilities are calibrated, its validation spatially honest, and its conclusions robust to the small sample sizes that inevitably characterise frontier exploration areas.

One of the most scientifically interesting findings concerns which predictors the models deemed important. Lineament density—a measure of how densely fractured and faulted the bedrock is, as expressed in the magnetic data—ranked as the single most influential predictor in both models. This is a direct quantitative endorsement of the structural-control hypothesis: orogenic gold in the Birimian is emplaced where crustal-scale shear zones and their subsidiary fractures channelled gold-bearing hydrothermal fluids into sites of pressure drop and chemical reaction. Yet when the predictors were grouped thematically, the structural-magnetic group and the hydrothermal-alteration group contributed comparably overall, suggesting that neither structure nor alteration alone tells the whole story. Gold accumulates where permeable structures intersect chemically reactive ground, and the models appear to have learned exactly that conjunction.

The practical payoff is captured in the maps’ ability to concentrate known gold into small, actionable areas. The high-prospectivity class of the random forest model captured 92.5 percent of the 67 known occurrences within just 159 square kilometres—22.52 percent of the study area. The decision tree’s high-prospectivity zone captured 80.6 percent of occurrences over 91 square kilometres, or 12.89 percent of the area. In exploration terms, this is the entire point of prospectivity mapping: a company or geological survey with finite drilling budgets can focus follow-up work on the delineated high-prospectivity zones, which the authors describe as providing a defensible basis for prioritising exploration at Beposo. The fact that the random forest achieves its higher capture rate over a larger area, while the decision tree is more aggressive in its discrimination, illustrates a genuine trade-off between sensitivity and specificity that explorers must weigh against their own risk tolerance.

Beyond the immediate results, the study’s methodological template may prove its most durable contribution. The workflow—derive a mineral-systems conceptual model, translate it into geophysically derived predictors, label occurrences with independent geochemistry, optimise hyperparameters under spatially independent validation, and evaluate with calibrated probabilistic metrics—offers a framework that is potentially transferable to analogous Birimian terranes across Ghana and the wider West African Craton, from the Sefwi belt to the greenstone belts of Burkina Faso and Côte d’Ivoire. The authors themselves note that this transferability remains a hypothesis to be tested on independent datasets, an appropriately cautious framing. Still, in a period when the global mining industry is scrambling to replace depleting reserves and machine learning is increasingly touted as the future of exploration, the Beposo study provides something rarer than hype: a carefully validated, statistically transparent demonstration that decades-old airborne geophysical surveys, a sound geological concept, and well-implemented classifiers can converge on the hidden architecture of gold mineralisation.

Subject of Research: Machine learning-based gold prospectivity mapping using airborne geophysical data in the Birimian Supergroup of Ghana

Article Title: Structurally controlled gold in the Beposo area, Ashanti Region, Ghana: testing a mineral-systems targeting model with random forest and decision tree classifiers

Article References: Forson, E. D., Nunoo, S., Amponsah, P. O., Osei, R. D., & Sakyi, P. A. (2026). Structurally controlled gold in the Beposo area, Ashanti Region, Ghana: testing a mineral-systems targeting model with random forest and decision tree classifiers. Earth Science Informatics, 19(10), Article 171. https://doi.org/10.1007/s12145-026-02221-8

Image Credits: AI Generated

DOI: 10.1007/s12145-026-02221-8

Keywords: gold, Ghana, Birimian Supergroup, machine learning, random forest, decision tree, mineral prospectivity mapping, airborne geophysics, orogenic gold, Ashanti Region, exploration targeting, structural geology