The Gold Standard of Archaeology | Electromagnetic Induction, Conductivity Mapping and Calibration Drift

1. Define the target and the expected physical contrast

Back to contents

Begin with the archaeological question, target size and depth, soil, geology, moisture, salinity, vegetation, modern metal and land use. A filled ditch may conduct differently from surrounding subsoil; fired material or metal can affect the in-phase response. These are conditional expectations, not material labels. Decide whether conductivity, magnetic response or both can answer the question, and state what a blank map would fail to exclude.

2. Understand primary and secondary fields

Back to contents

The transmitter coil creates a primary electromagnetic field. That field induces currents in conductive ground, and those currents create a secondary field detected by the receiver. Instruments convert phase relationships and amplitude into apparent conductivity or related responses under simplifying assumptions. The measurement integrates a sensitivity volume; it does not sample one point at one depth. Strong magnetic or metallic effects can also violate the assumptions used for conductivity conversion.

3. Choose coil orientation, spacing and frequency deliberately

Back to contents

Coil separation, orientation, operating frequency and instrument height control the depth-weighting and footprint of the response. Horizontal and vertical coplanar arrangements emphasise different volumes, while multi-receiver or multi-frequency systems provide several overlapping sensitivities. Greater nominal depth does not guarantee sharper deep archaeology. Record the exact configuration, direction of travel and sensor position so patterns can be compared and forward-modelled.

4. Control height, tilt and positional error

Back to contents

Because response changes with sensor height and orientation, uneven ground, operator motion and vegetation can create striping or false gradients. Use a stable sled or carriage where appropriate, log sensor height and tilt, and align positioning timestamps with instrument readings. Turning at line ends can alter orientation and speed. A dense grid with poor positional synchronisation can look detailed while placing anomalies in the wrong location.

5. Treat calibration and drift as survey variables

Back to contents

Instrument zero, temperature, electronics and warm-up can shift readings through the day. Cross-calibrate channels and configurations at a stable reference area or against an independent method when quantitative comparison matters. Repeat tie lines and control points at planned intervals. Correction should follow recorded controls, not a polynomial chosen because it flattens an inconvenient archaeological trend. Preserve both raw and corrected values and state every adjustment.

6. Worked example: a courtyard mapped morning and afternoon

Back to contents

In an invented 60-by-40-metre courtyard, morning EMI lines show a low-conductivity rectangle and a gradual eastward rise. Afternoon lines repeat the rectangle but the background is uniformly six millisiemens per metre higher. A repeated control line reveals the same shift, and sensor temperature rose by twelve degrees. The team applies a control-based offset correction, not a spatial detrend, and the rectangle persists across perpendicular traverses. A modern cable produces a sharp in-phase anomaly nearby. The result supports a conductivity contrast, while material identity awaits coring or excavation.

7. Process maps without manufacturing continuity

Back to contents

Inspect raw profiles, line direction, timestamps, speed, height, temperature, control repeats and channel agreement before interpolation. Remove spikes only under a declared rule and retain flagged data. Destriping can suppress genuine features aligned with traverses; smoothing can join separate anomalies. Display common scales when comparing configurations or dates, and show profiles across important patterns so the interpolation surface does not become the only evidence.

8. Changed-context transfer check: saline clay and corrugated fencing

Back to contents

Move from a quiet loam courtyard to saline clay beside a metal fence. The same workflow defines targets, logs configuration and repeats controls. The context compresses conductivity contrast in a highly conductive background, while the fence distorts the electromagnetic field and can dominate the in-phase channel. Reorient or offset the grid, model a buffer and use resistance, magnetometry or excavation where appropriate. Calibration cannot create target contrast that the ground and geometry never supplied.

9. Know the limits and common misconceptions

Back to contents

Apparent conductivity is not a direct soil sample, exact depth or unique material identification. Depth of exploration depends on configuration and conductivity structure rather than one fixed number in a brochure. Non-contact operation does not make the method immune to motion, height or metal. A zeroed instrument can still drift later. Conductivity and magnetic response may mix outside low-induction assumptions, and visually smooth maps can conceal configuration-dependent artefacts.

10. Audit acquisition, correction and validation

Back to contents

Archive instrument model and firmware, coil geometry, frequencies, sampling rate, sensor height, orientation, GNSS or total-station data, warm-up, temperatures, reference checks, tie lines, raw channels, corrections and interpolation parameters. Retain anomalies attributed to modern services as well as archaeological candidates. Link cores, trenches and companion geophysics to predictions recorded before validation, including contrasts that were rejected.

11. Frequently asked questions

Back to contents

Does EMI need electrodes? No; coupling is electromagnetic. Is apparent conductivity a true layer value? No; it is a weighted response. Can one configuration give exact depth? No. Why repeat control lines? They reveal drift and positional inconsistency. Does a high-conductivity anomaly mean a ditch? Not necessarily; moisture, clay, salinity or services can produce it. Should every broad trend be removed? No; it may be geology or archaeology.

12. The gold standard keeps the instrument inside the interpretation

Back to contents

Credible EMI mapping begins with target physics and a configuration matched to the expected scale. It stabilises height and position, measures drift through independent repeats, processes conservatively and checks important patterns across channels or methods. The best map is not the smoothest conductivity surface. It is a traceable measurement in which archaeological contrast remains visible after the instrument’s own movement, temperature and geometry have been accounted for.