Landslide monitoring Drone Guide

# Landslide Monitoring Drone Guide – Forestry

Introduction

Landslides are a significant risk in forested and mountainous environments. Heavy rainfall, flooding, erosion, wildfire, road construction, timber harvesting, geological conditions and changes in vegetation can all contribute to slope instability.

For forestry organisations, the consequences can extend far beyond the movement of soil and rock. Landslides can block forest roads, damage bridges and drainage systems, affect utilities, interrupt timber operations, restrict emergency access and introduce sediment into rivers and streams.

Traditional landslide assessment relies heavily on geologists, geotechnical engineers, surveyors and ground instrumentation. These remain essential, particularly when determining whether a slope is stable or predicting how it may behave.

Drones provide an additional monitoring capability.

Photogrammetry can create detailed 3D models of slopes. LiDAR can provide terrain information beneath parts of the forest canopy. RGB cameras can document visible cracks, erosion, exposed soil and changes to vegetation. Repeat surveys can then be compared to identify movement or changes in terrain.

This makes drones particularly valuable for rapid reconnaissance, repeatable mapping and directing specialist ground investigation toward areas showing potential change.

A drone should not independently declare a slope safe or predict that a landslide will occur. Its greatest value is providing high-resolution evidence that geotechnical and forestry professionals can incorporate into a wider monitoring programme.

Understanding Landslide Risk in Forest Environments

Forest terrain can experience several forms of slope movement, ranging from small shallow failures to much larger landslides involving substantial volumes of material.

Water is frequently an important factor. Prolonged rainfall can increase soil moisture and groundwater pressure, while concentrated runoff may erode slopes and road embankments.

Forestry infrastructure can also influence drainage. Roads, ditches and culverts alter the way water moves through the landscape. A blocked culvert may redirect large quantities of water toward a vulnerable slope.

Vegetation adds another layer of complexity.

Roots can contribute to soil stability in some environments, while harvesting may alter vegetation cover, drainage and surface conditions. Wildfire can remove vegetation and change soil characteristics, potentially increasing erosion and post-fire landslide risk.

Because these processes interact, landslide monitoring should consider the wider landscape rather than simply photographing the visible failure.

Drone mapping can help provide this broader geographic context.

Rapid Post-Storm and Post-Landslide Assessment

One of the strongest drone applications occurs immediately after severe rainfall, flooding or a reported slope failure.

Ground access may be unsafe or impossible.

A drone can provide an initial overview without requiring personnel to immediately enter unstable terrain.

RGB imagery can show the visible extent of the landslide, debris accumulation, affected roads and nearby infrastructure.

Aerial footage can also identify whether fallen trees, rock or soil have blocked access routes.

The upper part of the failure and the debris area below it can be documented from multiple perspectives.

This information can support the initial response.

Forestry managers may use the imagery to understand whether an alternative access route is required, while geotechnical professionals can identify locations requiring detailed investigation.

The drone provides reconnaissance rather than a final stability assessment.

Photogrammetry and 3D Landslide Mapping

Photogrammetry is particularly useful for documenting landslide geometry.

The drone captures overlapping images from multiple positions.

Processing software reconstructs the surface as a 3D point cloud or digital surface model.

This allows specialists to examine the shape and extent of the affected area.

Orthomosaics provide a detailed map, while 3D models allow the slope to be examined from different viewpoints.

Where suitable accuracy has been achieved, the dataset can support measurements of visible features.

Repeat photogrammetric surveys are especially valuable.

A slope can be mapped after an initial failure and then surveyed again following additional rainfall.

Differences between the models may indicate where material has moved.

Reliable change measurement requires consistent survey methodology, suitable georeferencing and an understanding of the uncertainty within each dataset.

LiDAR and Forested Slopes

Dense vegetation can make landslide monitoring difficult using RGB imagery alone.

Tree canopy may hide much of the ground surface.

LiDAR provides an important additional capability.

Laser pulses can produce returns from different levels within the vegetation. Some may reach the ground through gaps in the canopy.

Processing can classify these returns and create a digital terrain model.

This can reveal terrain features that are difficult to recognise from ordinary aerial photographs.

Scarps, depressions, drainage channels and previous landslide features may become more apparent.

Historical slope movement may also leave characteristic terrain patterns beneath vegetation.

LiDAR does not completely remove the effects of dense forest.

Ground-point density depends on vegetation, sensor characteristics, flight parameters and processing.

Nevertheless, it can significantly improve understanding of forest terrain.

Detecting Surface Cracks and Visible Warning Features

Surface cracks can sometimes accompany slope movement.

High-resolution drone imagery may document larger visible cracks where they are not obscured by vegetation.

Repeat flights can show whether these features appear to change.

Other visible indicators may include exposed soil, tilted vegetation, displaced road surfaces, erosion channels or changes in drainage.

These observations can be mapped within GIS.

A potential feature can then be assigned to a geotechnical team for investigation.

Caution is essential.

Not every surface crack indicates an active landslide, and a dangerous slope may not display clearly visible warning features.

Aerial imagery should therefore support professional assessment rather than become the sole basis for predicting failure.

Forest Roads and Slope Stability

Forest roads are closely connected to landslide management.

Road cuts may expose slopes, while embankments create artificial terrain.

Drainage systems associated with roads can also influence water movement.

Drones can inspect long sections of road relatively quickly.

Visible cracking, settlement, erosion, blocked drainage and slope failures may be identified.

Photogrammetry can provide detailed models of problem locations.

LiDAR may support broader terrain analysis.

After severe weather, forestry organisations can use drones to screen road networks before sending heavy vehicles into remote areas.

A road that appears visually intact should not automatically be considered structurally safe.

Subsurface movement or hidden instability may still exist.

Drainage, Culverts and Water Movement

Water management is central to many forestry landslide-monitoring programmes.

Drone imagery can help identify visible drainage conditions around vulnerable slopes.

Blocked ditches, overflowing culverts and concentrated runoff may be visible.

Photogrammetric or LiDAR-derived terrain models can help specialists understand potential surface flow paths.

GIS can combine this terrain information with mapped culverts, roads, streams and previous landslides.

This creates a wider picture of how water interacts with the landscape.

Following heavy rainfall, the same locations can be inspected repeatedly.

Buried drainage and groundwater conditions cannot normally be determined from ordinary aerial imagery.

Ground instrumentation and hydrological investigation may therefore be required.

Harvesting and Vegetation Change

Forestry operations can significantly change the appearance and hydrological behaviour of a landscape.

Pre- and post-harvest drone surveys provide a useful record.

Managers can document road construction, drainage changes and vegetation removal.

Multispectral imagery may provide supplementary information about vegetation condition.

LiDAR can quantify changes in canopy structure.

These datasets can help geotechnical and forestry specialists understand how land-management changes relate spatially to known slope hazards.

It would be inappropriate to assume from drone imagery alone that harvesting caused a particular landslide.

Slope failures usually involve multiple interacting factors.

Drone data provides evidence about changes in the landscape that can be incorporated into professional investigation.

Wildfire and Post-Fire Landslide Risk

Wildfire can significantly alter forest slopes.

Vegetation may be lost, soils can change and surface runoff may increase.

Following intense rainfall, burned landscapes can experience erosion, debris movement and slope failures.

Drones can map burned terrain before major rainfall occurs.

This provides a baseline.

Following storms, new surveys can identify visible erosion, debris flows or terrain changes.

LiDAR and photogrammetry can provide detailed terrain information.

These datasets can help specialists prioritise slopes for closer monitoring.

Drones can also inspect roads and drainage infrastructure following a wildfire.

This can be particularly important for maintaining emergency access.

Monitoring Movement Over Time

The greatest value of drone landslide monitoring often comes from repeatability.

A single survey provides a snapshot.

A sequence of surveys provides evidence of change.

The same slope can be mapped periodically or following significant rainfall.

3D models from different dates can be compared.

Areas showing measurable surface change can be highlighted.

This can help determine where additional instrumentation or ground investigation should be concentrated.

The frequency of surveys should reflect the risk and monitoring objective.

A relatively stable historical landslide may require occasional surveys, while an actively changing slope may require much more frequent observation and dedicated ground instrumentation.

Drones should be one component of this monitoring strategy.

Measuring Landslide Volume and Terrain Change

Where survey quality is suitable, 3D datasets can support estimates of terrain volume change.

A pre-event terrain model can be compared with a post-event model.

This may help estimate how much material has moved.

Deposition areas can also be mapped.

These calculations can support debris-removal planning and engineering assessment.

Accuracy depends on the quality of both datasets.

Vegetation can introduce additional uncertainty, particularly when comparing surface models rather than true terrain models.

Volume estimates should therefore include appropriate understanding of measurement uncertainty.

For important engineering decisions, professional survey and geotechnical standards should be followed.

AI and Automated Change Detection

Large forestry estates may contain many slopes requiring observation.

Manually comparing every survey can become time-consuming.

AI and automated change-detection software can help.

Current imagery or terrain models can be compared with historical datasets.

The software can highlight areas showing significant differences.

Computer vision may also identify visible road obstructions, exposed soil or other changes.

This allows specialists to focus their attention.

The AI system should not automatically declare that a landslide is imminent.

Terrain change may have many causes.

Forestry operations, road maintenance, vegetation changes and survey differences can all produce apparent changes.

Human interpretation remains essential.

GIS and Landslide Risk Mapping

Drone data becomes more valuable when integrated into GIS.

Terrain models can be combined with forestry roads, streams, culverts, buildings, utilities and harvesting compartments.

Historical landslides can be added.

Rainfall and environmental information may also be incorporated.

Each monitored slope can have its own digital record.

This might contain drone imagery, LiDAR data, previous inspections, photographs and geotechnical information.

Over time, the GIS becomes a landslide-management platform rather than simply a map.

Forestry managers can see which slopes may affect roads, infrastructure or waterways and prioritise monitoring accordingly.

Ground Sensors and Drone Integration

Drones provide excellent spatial information but are not continuously present.

Ground sensors can provide continuous measurements at selected high-risk locations.

Inclinometers, GNSS monitoring points, extensometers, piezometers and other geotechnical instruments may be used depending on the site.

The two approaches complement one another.

A sensor may indicate that movement has occurred at a specific location.

A drone survey can then map the surrounding slope.

Alternatively, drone change detection may identify an area where specialists decide additional instrumentation is required.

This creates a layered monitoring system.

Drone-in-a-Box and Automated Monitoring

Remote forestry areas may benefit from automated drone infrastructure.

A Drone-in-a-Box system could potentially be positioned near known landslide-prone terrain or important forest infrastructure.

Following significant rainfall, an authorised survey could automatically capture updated imagery.

The new dataset could be compared with previous surveys.

If significant visible change is detected, forestry and geotechnical teams could be notified for review.

Scheduled flights could also maintain a regular monitoring record.

Automated monitoring should not be confused with an automated landslide-warning system unless the complete system has been specifically designed and validated for that purpose.

The drone provides one source of information within the monitoring network.

BVLOS and Large-Area Slope Monitoring

Mountainous forestry estates may contain many kilometres of roads and slopes.

Long-endurance fixed-wing or VTOL drones can potentially survey much larger areas than short-range multirotors.

BVLOS operations may therefore support regional reconnaissance following major storms.

The aircraft could inspect road corridors and known landslide areas.

Specific locations could then be examined by smaller multirotors or ground teams.

BVLOS operations require appropriate regulatory approval, communications and risk controls.

Mountainous terrain introduces additional challenges because it can obstruct communications and create rapidly changing weather and wind conditions.

Reporting and Professional Interpretation

Landslide reports should clearly distinguish observations from engineering conclusions.

A drone report might state:

“New exposed soil and an approximately 12-metre surface feature are visible on the upper slope compared with the previous survey.”

This is an observation.

The report should avoid stating:

“The slope will fail.”

unless that conclusion has been made by appropriately qualified specialists using sufficient evidence.

Maps should show the date and expected accuracy of the survey.

Potential changes can be categorised for professional review.

This creates a defensible workflow in which the drone operator collects and documents evidence while geotechnical professionals interpret its significance.

Benefits, Challenges and Limitations

Drones can improve landslide monitoring by providing rapid access to terrain that may be difficult or dangerous to inspect from the ground.

Photogrammetry creates detailed 3D models.

LiDAR can provide valuable information about terrain beneath parts of the forest canopy.

Repeat surveys enable change detection.

RGB cameras document visible conditions, while GIS connects slope information with roads, streams and infrastructure.

The technology also reduces unnecessary exposure of personnel to unstable terrain.

However, significant limitations remain.

Vegetation can obscure the surface.

Weather can prevent flight.

GNSS and communications can be challenging in steep valleys.

Groundwater conditions are largely invisible to ordinary drone sensors.

Small subsurface movements may occur without obvious surface evidence.

A slope can also fail without displaying clear visual warning signs.

Drones should therefore complement, not replace, geotechnical investigation and ground instrumentation.

The Future of Forestry Landslide Monitoring

The future of landslide monitoring is likely to involve increasingly integrated sensor networks.

Satellites may identify broad areas of terrain change.

Drones can then provide higher-resolution mapping.

LiDAR can reveal detailed terrain beneath forest vegetation.

Ground instruments can continuously monitor movement and groundwater.

Weather stations can provide rainfall information.

AI can compare these datasets and identify locations where conditions appear to be changing.

A forestry organisation could maintain a digital terrain model covering its entire estate.

Known landslides and vulnerable slopes would be monitored over time.

Following extreme rainfall, Drone-in-a-Box systems or long-range drones could collect updated information.

AI would highlight significant differences for professional review.

Geotechnical specialists could then combine the aerial information with ground instrumentation and field observations.

The long-term direction is toward an integrated slope-monitoring platform in which drones provide high-resolution terrain data, LiDAR improves visibility beneath vegetation, satellites provide regional monitoring, ground instruments measure local movement and groundwater, weather sensors provide environmental context, AI identifies changes, and geotechnical professionals determine the actual level of risk.

Conclusion

Landslide monitoring is a valuable forestry application for drones because slope failures frequently occur in remote, steep and difficult-to-access environments.

The technology can provide rapid reconnaissance after storms, floods, wildfires and reported landslides.

Photogrammetry creates detailed 3D models. LiDAR provides valuable terrain information beneath parts of the forest canopy. RGB imagery documents visible cracking, erosion, debris and infrastructure damage.

The greatest value comes from repeated surveys.

Instead of viewing a landslide as a single event, forestry organisations can build a time series showing how the terrain changes.

When this information is combined with GIS, rainfall data, ground instrumentation and professional geotechnical assessment, drones can become an important component of a wider slope-monitoring programme.

They cannot predict every landslide or determine that a slope is safe.

Their role is to provide rapid, repeatable and geographically detailed evidence of visible terrain conditions and change.

Used appropriately, drones can help forestry organisations identify developing problems earlier, monitor vulnerable slopes more efficiently, reduce unnecessary exposure of personnel to unstable terrain and protect the roads, waterways and infrastructure that depend on stable forest landscapes.