Technology

How we see through rock, remotely.

Inside Earth's remote subsurface characterization combines three well-established signals — multi-spectral satellite imagery, a patented library of lab-calibrated signature materials, and Nuclear Magnetic Resonance interpretation — into a single workflow that maps hydrocarbons, minerals, lithium, and water up to 15,000 feet deep without field access, rigs, or environmental permits.

Request a Technical Briefing See Case Studies

15,000 ft

Max depth

30+

Calibrated materials

4–8 wks

Typical delivery

140+

Projects worldwide

The short version

Three signals. One answer: what's down there, and where?

Most geophysical methods image shape. Ours classifies substance. The difference matters because substance — is there oil, is there copper, is there lithium brine — is the economic question. Shape-based methods can locate anomalies that look right but hold nothing. Our output tells you, before a drill bit touches rock, whether the shape is filled with the thing you're after.

Layer 01 — Acquisition

Multi-spectral satellite imagery

The surface of the Earth carries signatures of what lies beneath it. Mineralogy at the surface, vegetation stress, thermal gradients, and subtle geochemical weathering patterns all correlate with subsurface composition. Most of these signatures are invisible to the human eye or to ordinary optical photography — but they are detectable across the visible, near-infrared, short-wave infrared, and thermal infrared bands that modern satellite platforms routinely capture.

We use publicly-available multi-spectral archives from Landsat, Sentinel-2, and commercial providers as our primary acquisition layer. For a typical project this returns terabytes of raster data per block — the overwhelming majority of which is noise, atmospheric interference, or surface land-use clutter.

The first step in our workflow is atmospheric correction, cloud/shadow masking, and terrain normalization. What remains is a corrected multi-spectral dataset across dozens of narrow bands — the raw material the rest of the pipeline consumes.

What satellites give us that traditional surveys don't

Coverage without mobilization. A license block of 100,000 hectares is routinely imaged in a single acquisition window. Field surveys at that scale take months.
Time-series depth. The public archives hold 40+ years of multi-spectral observations. Seasonal and inter-annual variation becomes a signal, not noise.
Cost structure. The imagery itself is either free (Landsat, Sentinel) or modestly priced. This lets us operate at a price point impossible for crew-based methods.

Earth from orbit — the first layer of our remote subsurface mapping pipeline

Layer 02 — Calibration

Patented lab-calibrated signature library

Raw multi-spectral imagery alone cannot tell you whether a basin contains oil, lithium brine, or copper sulphides. It gives you pixel values; you need to know which combinations of pixel values correspond to each target substance at depth.

That's what our signature library provides. It's an internally-developed corpus of lab-calibrated reference materials — physical samples of the substances we're looking for, plus their common host rocks and interfering materials — characterized under controlled conditions across the same spectral bands our satellite data captures. When we interpret a scene, we're comparing the observed signal to this reference library and returning the closest classified match per pixel, with a confidence score.

This library is the reason our method is general-purpose. We maintain signatures for:

Hydrocarbons — conventional oil, condensate, dry gas, and common contaminants (water, CO₂, H₂S).
Base metals — copper, lead, zinc, nickel in major sulphide and oxide host phases.
Precious metals — gold, silver, and platinum-group metals in their characteristic host rocks.
Battery and energy minerals — lithium (brine, spodumene, lepidolite), cobalt, vanadium, uranium.
Industrial minerals and rare earths — potash, phosphate, gypsum, selected REEs.
Hydrology targets — fresh groundwater, brackish brine, geothermal fluids.

The library is continuously extended. When a client engagement requires a material not yet characterized, our lab process can add a new signature in a matter of weeks.

Layer 03 — Interpretation

Nuclear Magnetic Resonance interpretation

Nuclear Magnetic Resonance — the same physics that drives MRI scanners and NMR well-logging tools — responds to the behaviour of atomic nuclei in a magnetic field. Different substances (oil, water, brine, specific minerals) produce characteristic NMR responses that distinguish them even when their other geophysical signatures overlap.

Our deployment of NMR is not downhole. It's an interpretation layer applied to the processed satellite-derived signals that the earlier pipeline stages produce — informed by the lab-calibrated signature library. The NMR step is what lets us say "this is hydrocarbon-saturated, that is water-saturated, this other zone is a gas cap" across an entire block, rather than just flagging a single "anomaly."

The output of this layer is a classified raster — every mappable pixel labeled with its most likely material type, with an associated confidence score. Downstream, these rasters are aggregated into the polygons, cross-sections, and coordinate lists the client receives as the deliverable.

Why NMR is the right interpretation tool here

It responds to substance, not structure — the exact question E&P and mining operators need answered before drilling.
It's well-established physics with decades of validation, including its downhole well-logging form routine in conventional plays.
It integrates naturally with the preceding spectral and signature layers — the workflow is mutually reinforcing rather than a leap of faith.

How the three layers combine

One workflow. Five steps.

01 — Scene acquisition

Multi-spectral imagery pulled and pre-processed for the target block. Atmospheric correction, cloud masking, terrain normalization.

02 — Signature matching

Each pixel compared against the lab-calibrated signature library; classified to the closest material match with a confidence score.

03 — NMR interpretation

Classification results re-weighted by NMR response characteristics to resolve ambiguous cases (e.g., oil vs water saturation).

04 — Calibration

Where ground-truth data exists (legacy wells, assays, drilling results, prior seismic), results are calibrated to match known outcomes in the region.

05 — Deliverable packaging

GIS-ready rasters, polygons, cross-sections, drill-target coordinates — formatted for the client's subsurface software of choice.

06 — Iteration

If the client returns with drilling outcomes, we close the loop — further refining the signature library and improving future deliverables in the region.

Depth & accuracy

What you can expect numerically.

Across 140+ commercial engagements we've tracked two operational metrics that matter commercially:

Hydrocarbon confirmation accuracy — false-positive rates in the single-digit percent range in well-calibrated basins with legacy wells. In frontier areas with no prior drilling, confidence scores per polygon are lower, but still sufficient for ranking prospects.
Depth-to-top-of-reservoir — typically within ±15% of subsequently drilled wells in mature basins. In offshore work with significant water columns, depth-to-top accuracy degrades proportionally to water-column attenuation.

Depth range: 0–15,000 feet (surface to target). For offshore work, this is combined water column + sub-seabed penetration — effective sub-seabed reach reduces in deeper water.

Lateral resolution: typically 30–100 metres per polygon depending on target, basin, and calibration data available. Higher-resolution interpretations are possible with narrow-band commercial imagery where the client wants to co-fund acquisition.

Every deliverable carries per-polygon confidence scores and explicit disclosure of the calibration data used. We are not a black box — we share methodology under NDA during scoping so technical evaluators can independently assess the fit for their specific target.

Go deeper, or compare against alternatives.

FAQ

Technical evaluators ask us.

Is NMR subsurface mapping the same as what seismic contractors call "NMR logging"?

No. NMR well logging is a downhole measurement made inside an already-drilled borehole. Inside Earth uses NMR as a remote signal-classification technique applied to surface-derived data — not downhole. The physics is the same; the deployment is fundamentally different.

How is this different from magnetotellurics, CSEM, or airborne gravity gradiometry?

Those methods image structural anomalies — contrasts in resistivity, density, or conductivity — that may or may not correlate with the target substance. Our method is tuned to the substance itself via a lab-calibrated signature library. We integrate MT, CSEM, and gravity data when legacy surveys exist.

Do you need ground access or any physical fieldwork to run a survey?

No. The entire workflow runs from satellite-derived data and our NMR signature library. No personnel visit the target area and no ground access permits are required for the survey itself.

How confident can you be in a remote-only result, compared to seismic + wells?

Every deliverable carries per-polygon confidence scores derived from our signal-strength metrics plus any available ground-truth calibration (legacy wells, assays, or seismic). We've delivered 140+ commercial projects with false-positive rates generally in the single-digit percent range in well-calibrated basins. Frontier areas without prior drilling have lower per-polygon confidence but are still useful for ranking.

Can your output be ingested into Petrel, Leapfrog, ArcGIS, or other subsurface software?

Yes. Deliverables ship as georeferenced rasters (GeoTIFF) and vector polygons (Shapefile / GeoJSON) in the client's project CRS. Standard imports into Petrel, Leapfrog, ArcGIS, Kingdom, and Micromine are supported. Custom formats on request.

Is this peer-reviewed, or proprietary black-box?

The underlying physics (remote spectroscopy, NMR signal classification) is well-established in open literature. Our patented contribution is the calibration library that maps measured signals to specific subsurface materials at depth, and the integration workflow. We share methodology under NDA during scoping so technical evaluators can independently assess fit.

Technical scoping

Want the full technical briefing under NDA?

Send us your basin or target. We'll return a project-specific scope, methodology disclosure, and indicative pricing within one business day.