Background
Magnetometry is one of the most reliable and widely used techniques in archaeological prospection, capable of identifying subsurface features based on variations in the Earth’s magnetic field. Ground-based surveys have provided archaeologists with highly detailed and precise maps of buried structures.
Recent advancements in drone technology have created opportunities to overcome some of the limitations of ground-based magnetic surveys. In particular, SPH Engineering’s system with SENSYS MagDrone R4 magnetometer has shown promise in drone-based archaeological prospection. By combining high-resolution magnetic data collection with the mobility and flexibility of drones, the research team decided to explore the effectiveness of this method at the Theilenhofen site.
Challenge
The team conducted this study with a drone-based approach to determine if drones could offer a faster, more flexible alternative to traditional ground-based magnetometry for surveying large archaeological sites. Ground-based methods, while highly precise, demand significant time and labor and are often restricted to easily accessible areas. Drones, in turn, can cover extensive areas quickly and access challenging or obstructed locations, making them ideal for large-scale initial surveys
Despite the potential of drone-based magnetometry, one of the primary challenges was the interference caused by the drone itself. Magnetic and mechanical noise generated by the motors and electronics of the UAV can degrade the quality of the magnetic data collected by the sensors. Minimizing these disturbances while flying at ultra-low altitudes, which is necessary for detailed and accurate magnetometric readings, poses a significant technical hurdle.
Additionally, the drone operates with True Terrain Following (TTF) technology, which enables it to fly extremely close to the ground — starting from just 40cm — while automatically adjusting to avoid collisions with the terrain. This ensures that the drone maintains a consistent ultra-low altitude relative to the uneven ground, essential for accurate magnetic measurements. Flight conditions such as wind and temperature fluctuations also play a crucial role in maintaining the stability and precision needed for these surveys.
Finally, there was the challenge of data processing. The noise introduced by the drone had to be removed through filtering, which involved the use of multiple methods, including low-pass filters, to isolate the archaeological signals from background noise.
Solution
Initially, a ground-based approach was used to establish a high-resolution reference dataset with a cesium magnetometer, well-known for its ability to detect fine archaeological details. The SENSYS MagDrone R4 was then tested in both ground-based and drone-based modes. In the ground-based setup, the R4 was mounted on a rack 0.3 m above the ground, with precise positioning provided by RTK GNSS.
To address the challenge of noise interference, the team mounted the SENSYS MagDrone R4 magnetometer system on a DJI M300 RTK drone, which also included RTK GNSS-based positioning for precise navigation. The magnetometer array was attached directly to the drone’s landing gear and operated at low altitudes, with the drone’s altitude monitored using the radar altimeter. This setup allowed the team to maintain the necessary flight altitude of 45 to 75 cm above the ground, a critical factor for collecting high-quality magnetic data.
To mitigate the magnetic interference caused by the drone, the researchers applied advanced data filtering techniques during post-processing. The SENSYS MagDrone DataTool software was used to remove noise signals, with filters designed to target specific frequency bands caused by the drone’s motors and electrical systems.
The team conducted surveys over a 3.8-hectare area of the Theilenhofen site, flying the drone at different speeds and altitudes to identify optimal conditions for archaeological feature detection. Data from the drone survey were then compared directly with the existing high-resolution ground-based magnetometer data, allowing the team to evaluate the effectiveness of the drone-based approach.
Outcome
The drone-based magnetometry survey provided highly encouraging results. The data collected by the magnetometer was able to detect significant archaeological features, including ditches, pits, fireplaces, and remnants of stone foundations. These are precisely the types of features that are critical in Roman archaeological prospection, as they reveal insights into the layout and function of ancient structures.
While the resolution of the drone-based data did not match the highest precision of the ground-based magnetometer, it was sufficient for identifying major archaeological features. The study demonstrated that drone surveys could rapidly cover large and difficult-to-access areas, providing a first-pass assessment that could guide more detailed, ground-based investigations.
Flying at ultra-low altitudes, combined with precise GNSS positioning and effective noise filtering, greatly improved the quality of the magnetic data. This allowed the team to establish best practices for future drone-based archaeological surveys.
Conclusion
The study conducted by Ludwig-Maximilians University highlights the growing potential of drone-based magnetometry in archaeological research. Although ground-based methods remain the gold standard for high-resolution data, drones offer a complementary approach that can rapidly survey large areas and identify key archaeological features. The ability to cover inaccessible or hard-to-reach sites is a significant advantage, and with further technological improvements, such as better noise reduction and enhanced flight controls, drone-based magnetometry may become an essential tool for archaeologists.
In the future, the integration of drone and ground-based methods may be considered the standard approach in archaeological prospection. Initial drone surveys can quickly identify areas of interest, which can then be followed up with detailed ground-based investigations.