Magnetometry Survey: Archaeology's Fastest Non-Invasive Tool
Magnetometry is the most widely used geophysical survey method in archaeology. It measures variations in the earth's magnetic field caused by buried features, producing maps that reveal ditches, pits, hearths, kilns, burnt structures, walls, and other archaeological deposits without breaking the surface. It is fast — a modern multi-sensor magnetometer system can survey several hectares per day — and it produces data that can be visualised in near-real time. Where the ground conditions are suitable, magnetometry produces some of the most striking archaeological images in the field: the buried plan of an entire Roman town, the outline of an unsuspected Bronze Age settlement, the circular ghost of a filled-in burial mound, all revealed in a pattern of magnetic anomalies across an open field.
The Physics
Every soil material has a magnetic susceptibility — a measure of how readily it is magnetised by an external field. The earth's own magnetic field magnetises soil to a degree proportional to its susceptibility. Archaeological features create magnetic anomalies (deviations from the background field) because their susceptibility differs from the surrounding undisturbed subsoil.
The most magnetically visible features are those that have been heated: when certain minerals in soil are heated above the Curie point (approximately 580°C for magnetite), they lose their existing magnetisation. On cooling, they realign with the ambient magnetic field and retain that alignment. The result is a strong thermoremanent magnetisation that makes heated features — kilns, hearths, burnt buildings, fire pits — the most reliable and strongest anomalies in magnetometry data. Pits and ditches filled with topsoil also show anomalies because topsoil has higher magnetic susceptibility than the subsoil horizon into which the feature was cut; bacteria living in the organic-rich fill enhance susceptibility through biochemical processes.
Walls and floors of fired brick or clay are strongly magnetic because they have been heated during manufacture or in use. Unburnt stone walls — particularly if built from low-susceptibility limestone or sandstone — may have lower susceptibility than the surrounding soil and thus produce negative anomalies (areas of lower-than-background magnetic field).
Instrumentation
The instrument most commonly used is the fluxgate gradiometer, which measures the vertical gradient of the magnetic field — the difference between readings from two vertically separated sensors. Gradient measurement cancels out regional variations and diurnal fluctuations in the earth's field, leaving only the local anomalies caused by near-surface features. A modern cart-mounted gradiometer system carries multiple sensor pairs (typically 5 to 10) spaced 0.5 metres apart, traversing the survey area in parallel transects. Data is logged at 0.1 metre intervals along each transect.
Caesium vapour and potassium magnetometers are more sensitive instruments capable of detecting weaker anomalies and producing higher-resolution data; they are used for sites where the magnetic signals are faint, such as stone-built sites in areas with low-susceptibility soils.
Landmark Projects
The Stonehenge Hidden Landscapes Project (2010–2014) used magnetometry (alongside GPR and other methods) to survey 12 square kilometres around Stonehenge and revealed at least 17 previously unknown monuments in the landscape. The data demonstrated that Stonehenge was embedded in one of the most densely monument-filled landscapes in prehistoric Britain — a finding that transformed the interpretive framework for the site.
At Portus, the harbour city of Imperial Rome on the Tyrrhenian coast south of Rome, systematic magnetometry survey by the British School at Rome revealed the street plan, warehouse complexes, and canal system of the harbour infrastructure in extraordinary detail, providing a framework for targeted excavation that has confirmed and extended the geophysical results over fifteen years of fieldwork.
In the Czech Republic, magnetometry survey of the La Tene oppidum at Bibracte identified previously unknown craft production areas, road networks, and ritual deposits that had not been detected in earlier excavations. The data provided a basis for planning a research excavation strategy across a site too large for comprehensive open-area excavation.
Limitations
Magnetometry is less effective in geologically complex areas where variable rock types create a noisy background signal. In urban areas, buried service infrastructure — iron pipes, cables, and modern construction debris — creates strong anomalies that obscure archaeological signals. In areas with heavy clay soils, the very high background susceptibility of the clay can reduce the relative contrast of archaeological features. Sites in high latitudes close to the magnetic pole require corrections for the steeper angle of inclination.
The method detects anomalies; it does not identify them. A circular positive anomaly in the data could be a pit, a ditch terminal, a buried post-hole, or a natural feature. Interpretation requires experience with the local geology, familiarity with the types of features expected in the relevant period and region, and ideally calibration against excavated contexts at the same site. Magnetometry data should normally be interpreted alongside other geophysical methods and desk-based research before any conclusions are drawn.
Integration with Other Methods
Magnetometry is most powerful when combined with other geophysical methods. GPR (ground-penetrating radar) provides depth information that magnetometry cannot supply and detects stone walls that magnetometry misses. Earth resistance survey (measuring the resistance of soil to an electrical current) is particularly good for detecting stone walls and floors and complements magnetometry's strength with heated features and organic pit fills. A multi-method approach, integrating magnetometry for speed and areal coverage with GPR and resistance for targeted depth information, is now standard practice on large research surveys.