Can drone surveying be used in forests? Explaining suitability and unsuitability in 6 points

Table of Contents

• Can drone surveying be used in forests?

• Is the target you want to measure the ground surface or the canopy?

• Tree density and seasonal conditions

• Required accuracy and approach to as-built conditions

• Flight safety and the local environment

• Compatibility with required deliverables

• Is it assumed that it will be combined with ground surveying?

• Strategies for Successfully Surveying Forests with Drones

• Summary

Can drone surveying be used in forests?

Drone surveying can be used in forests. However, if you treat it the same way as for flat development sites or urban areas with many paved surfaces, you will often not achieve the expected results. In short, drone surveying in forests is a task where situations in which it is suitable and situations in which it is not are clearly divided. If you introduce it while leaving this unclear, you are likely to end up with cases where, although you flew the drone, the ground surface was not captured, or where point clouds were created but cannot be used for design or quantity calculation.

The reason why difficulty increases in forests is that the abundance of trees makes what is visible from the air unlikely to match the information you really need on the ground. In typical drone surveying, images taken from above are used to recreate terrain and shapes. However, in forests the camera often captures not the ground itself but the canopy — the tops of leaves and branches. Since what is seen from above is not the ground surface, practically important information such as ground surface undulations, road shoulders, the edges of slopes, the width of access roads, and drainage flow becomes difficult to interpret as-is.

On the other hand, being a forest does not mean it is completely unsuitable. For purposes such as wanting to get an overview of a wide area in a short time, checking the overall condition of forest roads and work tracks, comparing changes before and after felling, or quickly grasping the outline of collapsed sites and areas of sediment runoff, drone surveying is extremely effective. This is because being able to capture the whole area from the air is itself of great value. It is particularly advantageous for initial assessment and recording at sites where slopes are difficult for people to walk and inspect, or where the area is so large that conventional methods alone would take too much time.

In short, drone surveying in forests is not a universal replacement; it’s a technique whose appropriate application must be judged according to the objective. What matters is not whether you can fly in a forest, but first deciding what you want to measure in the forest and to what extent. If you reverse that order, the data you obtain cannot be fully used in practice. From here, to determine whether drone surveying is suitable or unsuitable for forests, I will explain in order the six items that practitioners should look at first.

1. Is the target you want to measure the ground surface or the canopy?

When judging whether drone surveying is suitable in a forest, the first thing to confirm is what exactly you want to measure. This may seem like a very basic point, but it is the most easily overlooked in practice. In forestry work, the information desired varies by the field personnel. Some want to know the surface topography, some want to know the distribution of trees and trends in tree height. Others may simply want an overview of the positions of forest roads and work yards, the overall extent of planned harvesting areas, or the damage at disaster sites.

If what you want to know is the extent of the canopy, the overall appearance of the forest stand, the layout of roads and work corridors, the condition of landslides, or the spread of openings after logging, drone surveying is very well suited. Because you can grasp the whole picture from above, area-based information that is hard to capture just by walking the site can be organized in a short time. It also has high recordability, and if you acquire the same area at different times, comparing changes is easy.

However, if what you need to know is the elevation of the ground beneath trees, the ground surface that serves as the basis for earthwork volume calculations, the fine undulations required for drainage planning, the cross-sections used in design, or the ground elevations for as-built verification, caution is required. In photo-based drone surveys, the denser the trees, the less the ground is visible. Because unseen ground cannot be reproduced accurately, even if a 3D model appears to have been created, you must distinguish whether it follows the top of the tree canopy or is close to the actual ground surface.

If this distinction is left vague, major misunderstandings arise in the field. For example, when you obtain a colored 3D model or point cloud, you tend to think "the terrain has been captured." However, in reality, in forests it often only captures the shapes of the tops of trees. Drone surveying is suited to situations where you want to understand what is visible from above. Conversely, it is not suitable for situations where you need to reliably capture ground surfaces that are not visible from above using only drones. Simply establishing this distinction at the outset greatly improves the accuracy of on-site decisions.

2. Tree Density and Seasonal Conditions

Next to consider are the density of the forest and the season. Even within the same forest, the difficulty of drone surveying can vary greatly depending on how trees are spaced and how foliage is arranged. It is dangerous to lump these together and conclude "forests are difficult" or "you can do it even in forests." In practice, drone surveying is fully practical in areas that are close to sparse woodland, in post-harvest compartments, along access roads, where slopes are exposed, or along rivers—places where the ground can be easily observed from above.

On the other hand, in forests where branches and foliage are densely overlapping and the ground is almost invisible, photo-based drone surveying becomes difficult. Even if correspondences between images taken from above can be established, those matching points tend to be biased toward the surfaces of leaves and branches, so they are unlikely to produce the desired ground-surface model. Moreover, in forests, not only is the ground obscured, but large contrasts between light and dark often occur: valley-side shadows, deep canopy shadows, and directional lighting caused by slope aspect all affect the interpretability of the processing results.

Season is also important. In areas with many deciduous species, periods when leaves are sparse make the ground easier to see and increase the amount of information that can be obtained. Conversely, when foliage is dense, it may be good for capturing the forest canopy but unfavorable for reproducing the ground surface. In areas dominated by evergreen trees, the improvement due to season may be small. Therefore, when considering drone surveying in forests, you need to consider not only the site location but also the vegetation type and the time of year.

Also, at times when visibility conditions change—such as immediately after logging or after undergrowth has been cleared—results can vary greatly even at the same location. In other words, in forests, location information alone is not enough; what matters is the current condition of the site. This approach is suitable for sites where the ground surface is at least partially visible and where meaningful shapes can be extracted from aerial images. It is unsuitable for sites where foliage is dense and the ground is almost completely hidden, yet ground-surface information is required. Even looking at a few photos of the site makes this judgment considerably easier.

3. Required Accuracy and Considerations for the As-Built Condition

When using drone surveying in forests, it is essential to first determine what level of accuracy is required. A common mistake made by practitioners is to assume, in blanket fashion, that "drone surveying must be highly accurate." In reality, whether drone surveying is suitable for forest work depends greatly on the required level of accuracy.

For example, when it comes to getting an overall grasp of current conditions, scouting routes, checking the situation at disaster sites, or preparing materials to share with stakeholders, being able to visualize a wide area quickly is more important than small measurement errors. In such cases, drone surveying is sufficiently useful even in forests. That is because it is highly valuable for organizing the big picture in a short time. Even if there are some uncertainties in the finer details, it is often practical in the early stages of decision-making.

However, when it comes to applications that directly affect design—height control, cross-section creation, earthwork volume calculation, determining ground elevations for construction planning, and the basis for as-built management—the situation changes. In forests, because it is inherently difficult to accurately reconstruct ground surfaces that are not visible, the higher the required accuracy, the harder it becomes to rely on drones alone. What matters here is not only the magnitude of the error but whether the cause of the error can be explained. In forests, the fact that the ground is not visible itself becomes a source of error. In other words, even if processing settings are optimized, there is a limit when the source data lack ground-surface information.

Also, in forests you should be aware that confidence differs between horizontal and vertical directions. While it can be used to grasp planar positional relationships, care is needed when interpreting elevation. For example, you may be able to discern the alignment of forest roads, but you may not be able to determine the subtle height differences of road shoulders or the depth of side ditches. If you convert results directly into drawings without understanding this difference, you will face increased rework in later stages.

When using drone surveying in forests, it's important to clarify what decisions the data will be used for, rather than how many centimeters of accuracy you are aiming for. It is well suited to tasks where area-based visualization has high value, such as gaining an overall understanding, making comparisons, and checking general conditions. It is not suitable for work that intends to carry out precise height management based on an unseen ground surface without additional verification. In forests, the higher the accuracy requirements, the more important it becomes to secure supplementary observations and validation points.

4. Flight Safety and Local Environment

When considering drone surveying in forests, flight safety is as important as how easy it is to acquire data. If you judge only by whether something can be photographed from the air, you may run into operational problems on site. Forests contain many obstacles and, although they may appear to have ample space, they are in fact environments where factors that make flying difficult tend to overlap, such as securing takeoff and landing sites, maintaining distance from trees, wind turbulence caused by valley terrain, and reduced visibility.

Particularly problematic is the ease of takeoff and landing. It is not enough for the airspace above to be slightly open; you need a flat area where the aircraft can be safely raised and lowered, and space to keep distance from surrounding obstacles. In forests, the ground may be soft, there may be many branches or grass, and the terrain may be sloped, so pre- and post-flight operations are more constrained than you might expect. It is also important whether operators can control the aircraft from a stable posture and whether an access route can be secured.

Also, in forests there are situations where visual confirmation becomes difficult. The aircraft can easily move behind trees, and changes in terrain can suddenly cut off line of sight. Even if it can fly overhead, an environment in which the operator cannot continuously monitor the aircraft’s condition presents an operational risk. Furthermore, along valleys and ridgelines the wind flow may be inconsistent, and stability can vary at the same site depending on the time of day.

Flight in forests should be judged, before considering whether data acquisition will succeed, by whether the operation can be carried out safely and repeatedly. Suitable sites are those where takeoff and landing space can be secured, where the flight path is relatively clear, and where operators can operate without undue difficulty. Unsuitable sites are those with dense trees and poor visibility, with little room for takeoff and landing, where even a slight control error can lead to a collision risk. A site where measurements might be possible is not the same as a site that can be run reliably as part of regular work. In practice, the latter perspective is indispensable.

5. Compatibility with Required Deliverables

Whether drone surveying in a forest is suitable ultimately depends on what you require as the final deliverables. Even at the same site, conclusions about suitability change if the desired outputs differ. If you proceed with contracting or planning without clarifying this, a mismatch will arise between the data acquired and the deliverables.

Drone surveying is relatively well suited to forests for things like current aerial imagery, materials that organize broad positional relationships, before-and-after comparisons, purposes for grasping the extent of tree crowns, checking work routes and access roads, and compiling overviews of disaster or collapse sites. Because the way things appear from above has intrinsic value, the strengths of drone surveying tend to be evident in forests. They are also easy to use as materials for sharing on-site, and it is an advantage that they make it easy to convey the situation to stakeholders.

On the other hand, deliverables that place strong demands on ground-surface accuracy—such as topographic maps, cross sections, earthwork volume calculations, design overlays, and detailed slope checks—need to be considered carefully. In forested areas, the visible surface may not correspond to the true ground surface being sought, so even if a deliverable looks well presented, it can be weak as a practical basis. In particular, deliverables involving quantities will cause difficulties later if it is not made clear which surface is being used as the reference.

Also, suitability varies depending on the location within a forest. For example, areas where the ground surface is visible—such as around forest roads, clear-cut sites, or exposed slope faces—tend to yield relatively usable results. Conversely, in compartments with a closed canopy, although a three-dimensional appearance can be obtained, it is difficult to produce ground-surface outputs that can be used for design or quantity estimation. In other words, you cannot necessarily treat the entire forest uniformly as a single deliverable.

It is well suited to cases that require deliverables strong in grasping and sharing the current situation. It is not suited to cases where you try to produce, using only a drone, deliverables that assume an exact reproduction of the ground surface. In practice, you are less likely to fail if you determine the acquisition method by working backwards from the final deliverable. This reverse planning is especially important in forests.

6. Is it assumed that this will be combined with ground surveying?

If you want to bring drone surveying in forests into practical use, the final thing you should confirm is whether you can plan on combining it with ground surveying. This is the most realistic point that determines success or failure in forests. The idea of wanting to complete everything with drones alone is understandable, but in forests the cases where that is appropriate are limited. Rather, how you connect aerial area mapping with on-the-ground point verification will determine the quality of the results.

In forests, drones are well suited to quickly surveying wide areas. By contrast, ground surveying is suited to reliably capturing locations that are not visible, important control points, design-relevant heights, key cross-section points, boundaries, and areas around structures. When this division of roles is established, you can produce a practical, easy-to-use dataset even in forested areas. Conversely, if you try to have only drones do everything—ground surface, heights, and quantities—without considering this division of roles, you will run into problems.

For example, the approach of acquiring the entire site by drone while supplementing with on-the-ground surveys for the centerlines and branch points of forest roads, the shoulder and toe of slopes, areas around drainage facilities, key points of access routes, and sections where the ground surface is difficult to see is a method that pairs very well with forests. This way, you can achieve both the speed of area coverage and the reliability of point measurements. For site personnel, it also makes clear which information comes from the air and which from ground verification, making it easier to explain the results.

Also, the concept of control points and verification points is important. In forests, it is realistic to secure positional references by utilizing clearings and to connect the required coverage. You should not try to complete everything beneath the trees; instead, an approach that uses surrounding conditions to improve overall quality is needed. This method suits sites where drones and ground surveying are treated as complementary. It is unsuitable for sites that skip all ground verification and attempt to make decisions directly affecting design and quantities based solely on aerial imagery. The more successfully a site uses drone surveying in forests, the clearer the division of roles between air and ground actually becomes.

Approaches to Successfully Conducting Drone Surveys in Forests

Considering the six points above, what is necessary to succeed in drone surveying in forests is aligning the measurement objectives and methods, rather than relying on expectations of high-performance equipment. Projects that tend to fail in forests usually begin with the idea that "if we just fly, it'll work out somehow." Conversely, projects that are more likely to succeed decide in advance "which areas can be captured from the air and which should be collected on the ground."

First, it's important not to treat the entire forest uniformly. Even within the same project, conditions differ between open sections, densely wooded sections, along routes, along valleys, areas of tree-cutting, and around work yards. If you look at the site by section and simply separate places that are easy for drones to capture from those that are difficult to capture, the accuracy of the planning will improve. This results in a far more practical workflow than spending time trying to force things into shape during post-processing.

Next, work backwards from the final deliverable. The way you need to collect data changes depending on whether you want materials for sharing, progress records, something to use for design, or something to use for quantity takeoff. In forests, even if you produce visually appealing 3D data, that does not necessarily translate into practical deliverables. As the field person in charge, you are less likely to fail if you prioritize what decisions the data can support over its appearance.

Furthermore, it is important not to neglect verifying on-site conditions. In forests there are many obstacles that cannot be seen on maps. Takeoff and landing sites, ease of access for delivery, sightlines, terrain, wind, the positions of workers, and surrounding safety management—conditions that can only be determined on site directly affect the results. If only the shooting conditions are slightly poor, they can sometimes be compensated for in post-processing, but deficiencies in safety cannot be recovered afterward.

Above all, in forests, aiming to "complete everything with drones alone" is less effective than aiming to "collect the necessary and sufficient data without undue effort"; the latter leads to more consistent results. Aerial area data is extremely powerful, but there are many situations where it only becomes practical when supplemented by ground-based point data. Drone surveying in forests relies less on the glamour of aerial photography and more on the unglamorous planning skill of designing the entire field operation. Whether you can adopt this mindset greatly affects both judgments about suitability and the effectiveness of implementation.

Summary

Drone surveying can be used in forests. However, that does not mean that everything in a forest can be measured. It is suited to tasks where aerial surface information is valuable, such as gaining a general overview of a wide area, recording areas where the canopy and ground are visible, overall inspections of forest roads and work/access tracks, comparison of changes, and initial assessment of disaster sites. It is not suitable for work that attempts to precisely reproduce ground surfaces hidden by branches and leaves using only drones and to complete design, quantity estimation, and fine height control all at once.

The six key points for judgment are: what you want to measure, the tree density and the season, the level of accuracy required, whether it can be flown safely, what deliverables are needed, and whether it will be combined with ground surveying. By clarifying these six points, you can largely determine whether drone surveying is suitable for a forest project. In forest fieldwork, it is more important that the data obtained can be used for practical decision-making than simply whether flights are possible.

In sites like forests, where aerial assessment is combined with on-the-ground supplementary checks, a system that can efficiently establish positional references and verify key locations determines the quality of the outcomes. For example, if you want to link aerial survey data while capturing necessary points such as forest roads, work roads, slopes, and areas around yards, the iPhone-mounted GNSS high-precision positioning device LRTK is also effective. Broadly assess with drones and reliably secure the necessary points on the ground—this approach makes surveying in forests practical for field use.