Materials Testing

What are you Looking for?
+1 866 300 9012   contact@measur.ca
SHOPPING CART

No Products in the Cart
TOTAL:
$0.00
PROCEED TO CHECKOUT
VIEW CART
img

title

  • details
qty X
price

Mapping Subsurface Ice in Quebec Using
Drone-Based GPR and LiDAR

By: Dr. Kynan Hughson, Maria Zavarce, Lukas Wilgosh

 Case Study 

September 15, 2025                  8 minutes read

Companies

SPaRS-Lab (University of New Brunswick)

Drone & Sensors

DJI M350 RTK
DJI M3E
Zond Aero LF GPR
DJI Zenmuse L2 LiDAR

Location

Chic-Choc Mountains, Gaspé Peninsula, Quebec, Canada

Application

Rock Glacier Characterization & Climate Research

Rock glaciers are tongue-shaped landforms composed of rock, debris, and ice that flow downslope under their own weight. These features are critical indicators of permafrost conditions, glacial history, and climate variability. Understanding their internal composition and surface morphology is essential for predicting the impacts of a warming climate on alpine environments.

 

In Quebec’s Gaspé Peninsula, these features remain poorly understood due to remote locations, rugged terrain, and dense vegetation, which make ground surveys challenging and costly. Without subsurface data, questions about ice content and permafrost evolution remain unanswered.

 

To overcome these barriers, the Surface Processes and Remote Sensing Laboratory (SPaRS-Lab) at the University of New Brunswick developed a UAV-based approach combining LiDAR for surface mapping and GPR for subsurface characterization, enabling detailed analysis of both morphology and internal structure.

 

This combined methodology not only advances knowledge of rock glaciers in southern Canada but also offers a repeatable framework for monitoring climate-sensitive features in remote areas.

Project Overview

Field operations were conducted in July 2025 by a four-person team comprising faculty, graduate students, and research assistants. Accessing remote launch sites required 2–3 hour hikes along primitive trails, during which the team transported all research equipment. Students played a key role in flight planning, field documentation, visual monitoring, and environmental observations, including wildlife and weather conditions.

 

Flight plans, designed using satellite imagery, followed a staged approach based on complexity, beginning with photogrammetry, progressing to LiDAR, and concluding with GPR. Initial photogrammetry results were reviewed in the field, allowing real-time adjustments to subsequent flights to fill coverage gaps and optimize data collection efficiency.

Map of survey area divided into sections

The goals were to generate high-resolution surface models of selected rock glaciers, detect subsurface ice-rich zones, and integrate these datasets into a comprehensive 3D model for scientific analysis.

 

The team selected several potential rock glaciers in the Chic-Choc Mountains in collaboration with Parc national de la Gaspésie. Due to dense subarctic taiga, only ridgeline vantage points with clear line-of-sight were accessible for UAV deployment. The SPH Engineering Zond Aero LF drone-based GPR mounted on an DJI M350 RTK, DJI Zenmuse L2 high-accuracy LiDAR, and DJI M3E-based photogrammetry were identified as the most efficient and cost effective tools for carrying out this research.

"Without the aerial access capabilities of the remotely piloted aircraft, the type of data that was gathered simply could not have been collected."

Dr. Kynan Hughson

University of New Brunswick’s Department of Earth Sciences

Why the Zenmuse L2 & Zond Aero LF GPR on DJI M350 RTK?

The DJI Zenmuse L2 LiDAR and Zond Aero LF GPR, both deployed on the Matrice 350 RTK platform, were chosen for their ability to overcome the challenges of mapping remote, vegetation-covered rock glaciers in the Gaspé Peninsula.

The Zenmuse L2 delivers high-density point clouds (~450 pts/m²) with survey-grade precision, achieving 4 cm vertical and 5 cm horizontal accuracy at 150 m AGL through integrated IMU and RTK capabilities. Its small laser spot size (4×12 cm @100 m) and support for up to five returns enable accurate ground detection beneath dense vegetation, essential for generating detailed Digital Elevation Models (DEMs). With a detection range of up to 450 m, IP54-rated weather resistance, and coverage of up to 2.5 km² per flight, the L2 offers both safety and efficiency in rugged alpine terrain. Additionally, its 20 MP RGB camera supports colorized point clouds, while seamless integration with DJI Terra ensures a streamlined data processing workflow.

 

Complementing the LiDAR, the Zond Aero LF GPR provided critical subsurface imaging capabilities to detect ice-rich zones and internal structures without disturbing the terrain. Its modular antenna system (100, 150, and 300 MHz) allowed flexibility for varying penetration depths and resolutions, achieving depths of up to 12 meters under optimal conditions. Powered by Real Time Sampling (RTS) technology and high hardware stacking, the system produced low-noise, high-resolution radargrams suitable for detailed interpretation of permafrost conditions.

Weighing under 1.5 kg, it was well-suited for UAV deployment and leveraged the Matrice 350 RTK’s RTK positioning for centimeter-level geotagging. 
 

These sensors delivered a complementary dataset that combined high-resolution surface morphology with internal structural information.

Project Execution

At each site, surveys were carried out through a staged workflow to maximize data quality and coverage. On average, each survey included about one hour of photogrammetry, one hour of LiDAR, and one hour of GPR

Photogrammetry (DJI M3E): Collected high-resolution imagery for orthophotos and initial digital elevation models, achieving ~2 cm/pixel ground resolution.

 

LiDAR (DJI M350 RTK with Zenmuse L2): Generated dense point clouds (~450 pts/m²) capable of penetrating vegetation for accurate terrain modeling.

 

Ground-Penetrating Radar (Zond Aero LF): Flown at ~1 m altitude with terrain-following to produce axial and transverse cross-sections, mapping subsurface ice and structural features.

 

All flights were supported by an Emlid Reach RS3 GNSS base station, delivering real-time RTK corrections for centimeter-level georeferencing. Flight plans were refined on-site when early photogrammetry revealed coverage gaps, ensuring efficient data capture. Despite moderate winds (~9 m/s) and occasional light rain, both the M350 and M3E performed reliably, underscoring the resilience of the workflow under field conditions.

"Measur was excited to partner with the University of New Brunswick on this project. With our comprehensive portfolio of products, we were able to provide everything needed to complete this project including the drones, LIDAR payload, GPR, and GNSS base station. Our ability to deliver and train customers on these systems reflects our team’s dedication, expertise, and commitment to exceptional customer service."

Lukas Wilgosh

National Sales and Product Director - Drones & Geomatics at Measur

Data Processing Workflow

Orthophoto and Digital Elevation Model (DEM) generation from photogrammetry was primarily done using Open Drone Map, whereas LiDAR point cloud processing and DEM generation was completed using DJI Terra. Processing each dataset required approximately 10–18 hours.


GPR processing was handled in the Prism2 software package. In general, the GPR processing workflow was as follows:

  • Background removal filter to suppress system noise 
  • Initial gain adjustment to improve data visibility 
  • Velocity model estimation 
  • Horizontal high-pass filtering to reduce banding artifacts 
  • Band-pass filter to reduce high- and low-frequency noise 
  • Topographic corrections 
  • Migration to improve signal-to-clutter ratio 
  • Final gain adjustment to optimize data for interpretation and visualization

Ultimately, all data products were integrated in a GIS environment to facilitate synergistic interpretation.  

Highlighted in green are interpreted subsurface reflectors, which correspond to internal layers within the rock glacier, including zones of potential ice enrichment. Note how some layers are inclined against the grade of the slope (from right to left), this is a characteristic feature of rock glaciers suggesting that periglacial processes continue to shape this environment. The shaded circular area emphasizes a region of interest where strong reflectors crosscut disrupted layers suggesting concentrated ice or compositional variation. Overall, the dataset provides detailed insight into layering, internal structure, and potential ice-rich zones within the surveyed feature.

Findings

The generated LiDAR and photogrammetry shape models suggest a complex and episodic flow history for the investigated lobes. Thanks to LiDAR’s ability to penetrate vegetation, high resolution DEMs show classic rock glacier morphologies with well-developed ridges and furrows 
at two of three research sites. In particular, one site displayed GPR profiles that indicate complex and deformed layering consistent with the rock glacier model. GPR amplitudes at this site also show considerable zonation, potentially indicating the presence of ice-rich permafrost 
or even massive ice at depth. Regardless of the presence or lack of ground ice, the dataset is an important tool for characterizing the degrading permafrost environment in southern Canada.

Deliverables

  • High-resolution LiDAR point clouds and DEMs
  • Orthophotos, morphological maps, and surface structure maps
  • GPR radargrams and subsurface models
  • Interpreted maps of ground ice distribution and potential groundwater networks

All datasets will be openly published for scientific use and will contribute to upcoming peer-reviewed publications.

Technology and Performance

The combination of DJI M350 RTK and multi-sensor payloads was essential for this study. UAVs provided access to remote sites that were otherwise unreachable. LiDAR allowed surface mapping under vegetation, while GPR revealed internal ice structure without disturbing the terrain.

 

Measured resolutions included orthophotos at approximately 2.5 cm per pixel, DEMs at 3.5 cm per pixel, and GPR penetration up to 20 meters at 100 MHz. The system performed reliably across rocky, vegetated, and partially frozen terrain.

"The GPR Performance was admirable and the system worked in all use cases over rock, water, and 
organic rich terrain"

Dr. Kynan Hughson

University of New Brunswick’s Department of Earth Sciences

Conclusion

This project contributes to understanding permafrost distribution and climate history in southern Canada. The dataset supports future monitoring of rock glacier movement and degradation. More broadly, it demonstrates that UAV-based LiDAR and GPR can be applied to remote and hazardous alpine environments, opening opportunities for glacier research, environmental monitoring, and risk assessment.

 

SPaRS-Lab successfully combined UAV LiDAR and GPR to overcome access challenges and provide unprecedented insight into rock glacier morphology and subsurface ice in the Gaspé Peninsula. This approach establishes a scalable framework for monitoring climate-sensitive alpine features and contributes a high-resolution dataset for both scientific research and long-term environmental monitoring.

Get in Touch

Give us a call or drop by anytime, we will answer all enquiries within 24 hours on business days. We will be happy to answer your questions.

YOU MAY ALSO LIKE

BACK TO TOP