Key Takeaways
- In geophysical targeting, the raw geophysics (magnetics, gravity and EM survey acquisition) is still classical, mature science. AI enters at the interpretation stage.
- AI prospectivity-mapping platforms like VRIFY DORA fuse multiple geophysical layers with geochemistry and geology to rank exploration targets automatically.
- DORA uses a Random Forest model to assign a “prospectivity score” per grid cell after fusing about 24 derived rasters. It is a real, deployed product, not a research prototype.
- Competing platforms include GeoVista AI and OreFox, which similarly apply ML to multi-dataset target generation.
TL;DR
Geophysical targeting with AI keeps survey acquisition and processing as they are. You then load the processed grids, plus geochemical and geological layers, into a platform like VRIFY DORA or OreFox. A trained model scores every grid cell, and geologists refine the top-ranked areas into drill targets.
What Is Geophysical Targeting?
Geophysics measures physical properties of the subsurface, such as magnetism, density and electrical conductivity. Surveys produce grids and models that show where those properties change. Geophysical targeting is the step that turns those maps into a list of places worth drilling.
Traditionally, a geophysicist and a geologist overlay the maps and look for anomalies that fit a deposit model. For example, a porphyry copper target might combine a magnetic low from magnetite-destructive alteration with a chargeability high from disseminated sulphides. That interpretation takes skill and time, and different teams can rank the same anomalies differently. As a result, geophysical targeting is a natural place for machine learning to help.
How Do I Interpret Geophysics for Targeting With AI?
The process has three steps: prepare the geophysics, fuse it with other data, then review the ranked output.
Preparing geophysical data for targeting
Geophysics itself has not fundamentally changed with AI. Flying a magnetometer survey, running a gravity survey and processing the grids is still classical work, typically in Geosoft Oasis montaj or similar software. Our post on airborne geophysics covers that processing step. The goal here is clean, georeferenced grids in a common coordinate system.
Fusing layers in an AI targeting platform
Next, assemble the geophysics with your geochemical and geological layers and upload the stack. VRIFY DORA is the clearest example in production. Its makers describe it as the first AI-assisted mineral discovery platform. It processes input datasets into about 24 derived rasters tailored to the deposit type you are targeting. A Random Forest model then assigns each grid cell a prospectivity score.
OreFox takes a similar data-fusion approach with two linked AI systems. “Prospector AI” compares your data against known-deposit signatures. “Hunter AI” refines that comparison into a visual target map. GeoVista AI is a comparable entrant worth evaluating alongside these.
Reviewing the ranked geophysical targets
Finally, analysts refine the scored cells into priority target zones. Check each zone against the geological model, access and land tenure before committing drill metres. Geochemical support matters too, so compare the ranking with your geochemical proxies where they exist.
What to Expect From AI Geophysical Targeting
DORA has a publicized success. In one case, an independent DORA-generated target and a human exploration team’s own target converged on the same gold discovery at Southern Cross Gold’s Sunday Creek project. That is encouraging, but it is one data point, not a guarantee.
Most of the real work happens before the model runs. Getting geophysics, geochemistry and geological interpretation into one clean, fusable dataset takes time. Moreover, these models learn from known deposits, so unusual or deeply buried targets may score poorly. Newer sensing methods, such as muon tomography, are starting to add density information at depth that surface surveys miss.
FAQs
What is geophysical targeting in mineral exploration?
It is the process of turning geophysical survey data into a ranked list of places to drill. Geologists look for anomalies in magnetism, density or conductivity that fit a deposit model. AI platforms now automate much of that ranking.
How does AI improve geophysical targeting?
AI fuses many geophysical, geochemical and geological layers at once, which is hard to do by eye. A trained model scores every grid cell consistently. Geologists then spend their time reviewing the best targets instead of overlaying maps.
Which platforms offer AI geophysical targeting?
VRIFY DORA, OreFox and GeoVista AI all apply machine learning to multi-dataset target generation. DORA uses a Random Forest model over derived rasters. OreFox uses its Prospector AI and Hunter AI systems.
Do I need new geophysical surveys for AI targeting?
No. These platforms use standard processed grids, so existing surveys are often enough to start. New surveys help where coverage is sparse or resolution is too coarse for the target size.
Try It With Geocluster
Getting geophysics, geochemistry and geological interpretation into one clean dataset is most of the real work before any AI targeting tool can help. That multi-source geoscience data wrangling is what Geocluster is built for. Read how the Geocluster Research Harness works to learn more.
