Geophysical Survey Principles in Forensic Contexts

Ground-penetrating radar, magnetometers, resistivity meters, and conductivity sensors are very different instruments, but they share the same underlying logic. Each is sensitive to a particular physical property of the ground, and each detects a target only when that target's property differs from the surrounding material. Understanding which properties a grave or buried object is likely to disturb is the first step in choosing the right method.

• Density: disturbed grave fill is typically less compacted than undisturbed host soil, producing a slight density deficit. Microgravity surveys can detect this, though instrument precision requirements make them rare in forensic fieldwork.
• Magnetic susceptibility: topsoil is enriched in iron-bearing minerals relative to deeper subsoil. When a grave is dug and the spoil mixed during backfilling, susceptible topsoil moves downward, creating a susceptibility anomaly above background.
• Electrical resistivity: moisture and ion content control current flow. Decomposing organic material releases fluids and ions, lowering resistivity. The grave fill may also retain moisture differently from undisturbed ground, sustaining the anomaly through seasonal variation.
• Dielectric permittivity: the electromagnetic contrast that GPR detects. Air voids, body cavities, clothing, and grave boundaries each produce partial reflections where permittivity steps sharply. This is the basis of GPR's grave-detection capability.

Instrument sensitivity is not the binding constraint in most forensic geophysical surveys. The binding constraint is the signal-to-noise ratio: how large is the anomaly compared to the natural variation in the background measurement? If the background fluctuates by 5 nanoteslas and the grave produces a 3-nanotesla anomaly, the grave is invisible no matter how sensitive the instrument.

Noise sources fall into two groups. Geological noise comes from natural variation in soil mineralogy, rock fragments, and moisture gradients. Cultural noise comes from buried infrastructure: pipes, cables, reinforcement rods, rubble, and old structures. Both scale with depth in different ways. At shallow depths, geological noise often dominates. In urban environments, cultural noise typically dominates, which is why urban forensic searches are among the hardest to interpret.

Depth limits vary by method and soil. GPR at 250 MHz in dry sandy loam can image to 2-3 m. The same antenna in waterlogged clay may lose all coherent signal at 0.5 m. Magnetometry over a magnetically quiet substrate (chalk, limestone) can detect a disturbed topsoil pocket at 1.5-2 m. Over a basalt-rich soil with strong natural magnetic variation, even a large iron object may not rise above background. Matching the method to the terrain's physical character is the practitioner's core competency.

No geophysical method works in all soils. Each has preferred conditions and failure modes; ignoring soil type wastes fieldwork time and produces results whose limitations are not communicated to investigators.

In practice, soil type information comes from a desk-based assessment before fieldwork. Geological maps, borehole records, soil survey data, and a brief site walk with a hand auger or probe to characterise the stratigraphy all feed into the method-selection decision. The British Geological Survey (BGS) 1:10,000 and 1:50,000 bedrock and superficial-deposits maps are standard resources in England and Wales. Equivalent national surveys exist for most countries with active forensic practice.

Using a single geophysical method introduces systematic risk. A positive result may be a true grave or a false positive from a geological feature or buried infrastructure. A negative result may reflect a genuine absence of target or simply the method's inability to image that particular target in that particular soil. Both errors carry real costs: unnecessary excavation wastes resources and disturbs context; a missed target can leave a case unresolved.

Multi-method surveying reduces both error types by exploiting the different physical sensitivities of each method. A buried body will produce contrasts in several properties simultaneously. A gas pipe may appear in one dataset (resistivity, EM conductivity) but not in another (magnetometry if non-ferrous). An anomaly that appears at the same position in two or more independent datasets is a much stronger candidate for excavation. Conversely, an anomaly that appears in only one of three methods is treated with lower confidence unless there is a physical reason why only one method would detect the target.

Survey design translates the physical principles above into a practical field plan. The three main variables are grid spacing (traverse line separation and point spacing along a traverse), instrument settings (antenna frequency for GPR, time window, gain; sensitivity and dynamic range for magnetometers), and spatial coverage (which areas to survey and in what order).

1. Define the target size and depth
   The minimum target dimension relative to survey depth controls the grid spacing needed to avoid missing it. A single body burial is roughly 0.4 m wide. Nyquist-type sampling logic suggests traverse spacing should be no more than half the target width, so 0.2 m spacing for confident detection of a single burial. Wider spacing is acceptable for reconnaissance if a follow-up pass at closer intervals is planned when anomalies are found.
2. Set instrument parameters for the expected depth
   For GPR: higher frequency gives better resolution but shallower penetration. A 250 MHz antenna is a common forensic starting point, resolving targets to roughly 12 cm at 1 m depth in dry ground. If the target is known to be shallow (less than 0.5 m), a 500 MHz antenna improves resolution. The time window must extend past the expected maximum depth, with margin.
3. Establish a site datum and grid
   All survey data must be georeferenced to a fixed datum so that multi-method datasets can be co-registered and anomaly positions reported in coordinates that excavators can find. A total station or RTK-GPS establishes the datum. The grid lines are marked with stakes or cord at the specified spacing before scanning begins.
4. Document all cultural features before scanning
   Drain runs, utility markers, fence posts, and any areas of obvious ground disturbance are mapped on the site plan before scanning. This protects against misinterpreting known infrastructure as a forensic target and ensures that anomalies in the final report can be checked against the pre-survey map.

Coverage strategy depends on prior intelligence. When a search area is large but the suspected location is partially constrained by witness information, a phased approach works well: a wide, coarser-grid reconnaissance pass over the whole area is followed by a denser pass over candidate zones. Prioritising sections with accessible terrain before completing the full area often finds the target faster in practice than a rigid left-to-right grid.

The geological context of a search area interacts with both the detection physics and the noise environment. On chalk downland (southern England, northern France, Cretaceous plains of North America), the substrate is magnetically quiet and electrically resistive. Magnetometry and GPR both perform well. On basalt-dominated terrain (parts of Scotland, Iceland, the Indian Deccan), the natural magnetic variation swamps small susceptibility anomalies from grave fill, making magnetometry unreliable unless the target contains significant metalwork. Resistivity and GPR become the preferred methods there.

River floodplains and coastal environments introduce a different challenge: high soil moisture and often elevated salinity. Saline pore water makes resistivity and EM conductivity signals large everywhere, masking the target contribution. GPR signal is heavily attenuated. In these settings, the magnetometer or a cadaver dog may be the only practical non-invasive tool before committing to systematic probing.

The practical framework is to read soil and geology data before any instrument is deployed, select methods that image properties likely to contrast in that specific terrain, and document the selection rationale in the survey report. That documentation matters in court: a practitioner who can explain why a given method was chosen for a specific terrain, and what the result does and does not mean, will withstand cross-examination. One who cannot articulate the terrain-method rationale will not.