Vegetation penetration surveys Drone Guide

Vegetation penetration surveys are one of the most valuable archaeological applications for drones because many historic landscapes, structures and earthworks are hidden beneath dense woodland, scrub or other vegetation. From normal aerial photography, these sites may appear to be little more than trees and undergrowth, but LiDAR-equipped drones can reveal the shape of the ground beneath much of that vegetation. This makes drone LiDAR particularly useful for archaeological prospection in forests and other difficult terrain. The laser scanner sends large numbers of pulses towards the ground, and some of those pulses pass through gaps between leaves and branches. By separating reflections from vegetation from those representing the underlying terrain, surveyors can generate a digital model of the ground surface. The resulting terrain model may reveal ancient roads, building foundations, defensive banks, burial mounds, terraces, field boundaries, mining features and other archaeological structures that are extremely difficult to identify from the ground. Drone vegetation penetration does not literally see through solid vegetation in the same way as X-ray imaging. Its effectiveness depends on how many laser pulses are able to reach the ground. Nevertheless, when conditions are suitable, drone LiDAR can transform archaeological survey by allowing researchers to examine large wooded landscapes in remarkable detail. ## **What Is a Vegetation Penetration Survey?** A vegetation penetration survey uses LiDAR or another suitable remote-sensing technique to map the terrain beneath vegetation. For archaeology, LiDAR is by far the most important technology. A drone carries a laser scanner that emits pulses towards the landscape while an accurate GNSS and inertial navigation system records the aircraft’s position and orientation. Each laser pulse may produce one or several returns. The first return may come from the top of a tree, another from branches and another from the ground below. Processing software classifies these points so that vegetation can be digitally removed and the underlying terrain can be analysed. ## **Why Archaeologists Use Drone LiDAR** Many archaeological sites are difficult to discover because vegetation hides their surface features. A stone wall that once stood several metres high may now survive only as a slight ridge. An ancient road may appear as a shallow depression. A settlement boundary may be represented by a subtle bank only a few centimetres or tens of centimetres above surrounding terrain. Inside dense woodland, these features can be almost impossible to recognise visually. LiDAR allows archaeologists to analyse the shape of the landscape itself rather than relying entirely on visible surface appearance. ## **How LiDAR Penetrates Vegetation** The term vegetation penetration can sometimes be misunderstood. The laser does not normally pass directly through leaves or solid branches. Instead, thousands or millions of laser pulses are transmitted across the survey area. Some pulses encounter vegetation, while others travel through small gaps in the canopy and reach lower vegetation or the ground. Modern LiDAR systems record these multiple reflections. When enough ground returns are collected, software can reconstruct the surface beneath the vegetation. ## **Multiple Returns** A single LiDAR pulse may produce several returns. The first return may represent the tree canopy. Intermediate returns can represent branches or bushes, while the final return may come from the ground. This ability to record multiple returns is particularly useful for forestry and archaeological mapping. The number and quality of ground returns influence how accurately the terrain beneath vegetation can be reconstructed. ## **Point Clouds** The raw output from a LiDAR survey is normally a point cloud containing millions of three-dimensional measurement points. Each point has X, Y and Z coordinates and may also contain information such as return number, intensity or classification. At first, the point cloud includes everything that the laser measured: trees, bushes, buildings, rocks and terrain. Processing is then used to separate these different features. ## **Ground Classification** Ground classification is one of the most important stages in an archaeological vegetation-penetration survey. Algorithms identify points believed to represent the actual ground. Vegetation points are removed from the terrain model. If the classification is too aggressive, archaeological features such as walls or mounds can accidentally be removed. If it is too conservative, vegetation may remain within the model. For archaeology, careful processing is therefore particularly important. ## **Digital Terrain Models** After ground points have been identified, they can be converted into a Digital Terrain Model, or DTM. The DTM represents the shape of the ground without most vegetation. Archaeologists can then analyse this model using different lighting and terrain-visualisation techniques. Subtle archaeological features can become much easier to recognise. ## **Digital Surface Models** A Digital Surface Model, or DSM, is different because it represents the highest visible surface. In woodland, that normally means the tree canopy. A DSM is useful for understanding vegetation structure and landscape context but does not expose archaeological terrain beneath the trees. Archaeological LiDAR projects often create both DSM and DTM products. ## **Hillshade Models** Hillshade is one of the most common ways to visualise LiDAR terrain. Software simulates sunlight shining across the landscape from a chosen direction. Small banks, ditches and terraces create shadows that make them easier to see. Changing the direction of the simulated light can reveal features that were almost invisible in another hillshade. ## **Multi-Directional Hillshade** Using only one artificial light direction can hide archaeological features that align with the illumination. Multi-directional hillshade combines several light directions. This can make linear features such as roads, banks and field systems easier to detect regardless of orientation. It is particularly useful for initial archaeological interpretation. ## **Local Relief Models** Local Relief Models help emphasise small variations in terrain while reducing larger-scale slopes. This is useful when archaeological features sit on hillsides. A normal elevation model may be dominated by the natural slope, while a Local Relief Model can make smaller banks and depressions stand out. This technique is widely useful for archaeological LiDAR interpretation. ## **Sky-View Factor** Sky-View Factor is another terrain-visualisation method. It estimates how much of the sky would be visible from each terrain point. Small ridges and depressions influence the value. This can reveal subtle archaeological features that are difficult to identify using traditional hillshade alone. ## **Openness** Positive and negative openness can also enhance terrain shape. These visualisations help highlight elevated and depressed features. Banks, ditches and pits can therefore become easier to recognise. Archaeologists often examine several visualisation methods rather than relying on a single map. ## **Ancient Settlements** One of the strongest applications is locating previously unknown settlements. Building platforms, enclosures, roads and defensive earthworks may survive beneath forest vegetation. LiDAR can reveal the overall settlement pattern. Researchers can then target selected areas for field survey or excavation. ## **Ancient Roads** Historic and prehistoric roads can survive as shallow hollow ways, raised banks or linear terraces. From ground level, they may be difficult to follow through dense woodland. LiDAR can reveal a continuous route across large areas. This helps archaeologists understand how settlements and landscapes were connected. ## **Roman Roads** Roman roads are particularly