GPR feasibility is site-specific. Conductive ground, moisture, reinforcement, congestion, surface coupling and target geometry can reduce penetration or obscure responses. A survey plan should state what success and inconclusive evidence would look like.
1. Define the decision
State the decision the survey must support: locate reinforcement before drilling, map a pavement interface, screen for local support variation, identify a buried feature, or select locations for physical verification. The objective controls the required resolution, depth range, positioning accuracy, line spacing and evidence standard.
Create a target model describing expected depth, dimensions, orientation, material contrast and plausible alternatives. GPR responds to changes in electromagnetic properties. A target without sufficient contrast, or one smaller than the practical resolution, may not be detectable.
2. Review constraints before mobilising
- Available drawings, previous surveys, cores, boreholes and repair history.
- Surface material, roughness, water, slopes, kerbs and inaccessible zones.
- Reinforcement density, metallic obstructions, vehicles and radio-frequency interference.
- Traffic, production, hygiene, permits and exclusion requirements.
- Known or suspected utilities and the safe-dig procedure.
- Survey control and references that will remain identifiable after the fieldwork.
A desk study should identify where GPR is likely to add evidence and where complementary methods or direct inspection will be required.
3. Select the system as a trade-off
Higher-frequency antennas can provide finer detail over a shallower practical range; lower-frequency antennas may retain broader or later responses with reduced detail. Actual performance depends on the complete system and site conditions. Frequency alone is not a penetration guarantee.
Ground-coupled systems support detailed local surveys but require suitable access and coupling. Air-coupled systems can acquire pavement information efficiently at speed, although their purpose and calibration differ. Multi-channel or stepped-frequency systems can improve coverage, but do not remove the need for traceable geometry and interpretation controls.
4. Design geometry around the target
Record each line’s origin, direction, length and spacing in a coordinate convention tied to permanent site references. Line spacing must relate to target size and orientation. Parallel lines can show repeatability; intersecting cross-lines are needed when the decision depends on lateral continuity or plan geometry.
Do not claim a closed anomaly boundary when cross-lines do not intersect the strongest response or when inaccessible strips break the grid. Distinguish the planned grid, surveyed coverage and unsurveyed areas.
5. Establish acquisition settings
Document antenna or band, trace spacing, samples per trace, sample interval, recorded time window, stacking, filters applied during acquisition, wheel calibration and positioning method. Set the time window to cover the target range with margin, while recognising that recorded time is not depth.
Trace spacing should be sufficient to sample the anticipated response. Overly sparse traces or lines cannot be repaired later by visual interpolation. Mark line direction consistently so distance coordinates can be reconstructed.
6. Use field quality controls
- Run a repeat line to test positional and response repeatability.
- Cross a known feature where available.
- Review early profiles before completing the grid.
- Check time-zero stability, clipping, saturation, coupling and encoder behaviour.
- Record shortened, obstructed, reversed or repeated lines immediately.
- Preserve raw files and field notes before processing or export.
Field adaptation is legitimate when documented. Silent changes to spacing, direction or settings weaken traceability.
7. Specify the deliverables
Require raw and processed data, file inventory, survey plan, coordinate convention, acquisition settings, processing record, representative profiles, interpreted findings, limitations and recommended verification. Images without source identifiers and axes are communication aids, not an auditable dataset.
Planning checklist
- What decision must the evidence inform?
- What target and alternatives are plausible?
- Is the expected contrast and geometry detectable?
- What coverage is required to test continuity?
- How will every line be positioned and reconstructed?
- Which known feature or repeat line will provide quality control?
- What conditions would make the result inconclusive?
- What follow-up is available if uncertainty remains?
Official reference
USACE EM 1110-1-1802 — Geophysical Exploration for Engineering and Environmental Investigations