As the Arctic climate warms, its permafrost landscapes are rapidly melting, causing significant changes to the terrain. High-resolution aerial photography has been an ideal tool for analyzing variations to the terrain in long-term investigations of permafrost thawing dynamics. To achieve improved image quality -- and ultimately help better gauge the impact of climate change on permafrost landscapes -- the DLR Institute of Optical Sensor Systems developed its Modular Aerial Camera System (MACS) Polar 18, designed for highly accurate 3D-reconstruction and monitoring of arctic terrain under extreme conditions.
MACS Polar 18 featuring the computing unit on left and sensor unit on right featuring SVS-Vistek HR Series RGB and NIR cameras. Image Credit: courtesy of German Aerospace Center and Potsdam University
Unlike conventional aerial photography, MACS Polar 18 acquires datasets in four-band (R-G-B-NIR) multi-spectral orthomosaics, 3D point clouds, and digital surface models. Multiple integration times per scene can be acquired to avoid under- or overexposed pixels where very dark areas, such as water and bare soil, and very bright surfaces, like snow and ice, coexist in targeted ground areas. The Polar 18 version of MACS was specifically adapted to operate reliably in harsh arctic areas subject to very low temperatures and difficult lighting conditions.
MACS Sensor Unit
The MACS Polar 18 sensor unit features three SVS-Vistek HR Series GigE industrial cameras with 16 megapixel resolution: two RGB cameras with overlapping right- and left-looking view directions, and one camera in the near-infrared (NIR) wavelength. The cameras are housed in a sensor unit along with an inertial measurement unit detecting acceleration, rotation and velocity. Connected to the sensor unit is a primary computer for processing and a global navigation satellite system (GNSS) receiver.
Arctic Exploration
During the summers of 2018, 2019 and 2021, scientists at the German Aerospace Center (Berlin, Germany) and Potsdam University (Potsdam, Germany) deployed the MACS Polar 18 to capture tens of thousands of images in a series of aerial campaigns with final results published in 2024. Flights were conducted over 1500 square kilometers of permafrost-affected landscapes in northwestern Canada and northern and northwestern Alaska. MACS was installed aboard two Basler BT-67 planes refitted for harsh polar environments with the sensor unit in a belly port of the planes. The planes flew in grid patterns which were suitable for photogrammetric processing.
The two SVS-Vistek RGB cameras and the IR camera were set to acquire 4864 x 3232 pixel resolution images at a continuous rate of 4 frames per second, and were electrically triggered to start image exposure at the precise same time. Camera parameters were fixed except for the exposure time to minimize motion-induced blur in flight. The cameras also delivered an electric pulse to the GNSS receiver, generating information documenting the exact time, position and attitude of the acquisition. That data was written on the raw image file. Cameras were only turned off during takeoff and landing, when low-level clouds occurred locally, over larger water bodies or sensitive infrastructure, during sharp turns, or when space on the hard drives for data storage was running low.
Upon acquisition, the raw Bayer-pattern images were stored in the proprietary MACS format. Multiple software packages were used in the workflow, but most data handling for pre- and post-processing was implemented automatically via Python scripts and WhiteboxTools. Additionally, images were processed using the Pix4Dmapper mosaic tool, which combined several images with overlapping areas to produce one seamless image. Orthorectification of each mosaic was automatically carried out in Pix4Dmapper with the internally calculated elevation information. Tiling the mosaics created large 3D maps where permafrost melting could be visually observed over a multi-year period.
Future Applications
In this study and others, MACS demonstrated the potential to open up new possibilities for data analysis as well as for the discovery and validation of various landforms. This includes tracking coastal erosion; analysis of ice-wedge polygons; monitoring thaw subsidence to evaluate potential impacts on infrastructure; the detailed analysis of recent and historic fire scars; the detection of lake drainages; the examination of individual shrubs and trees in the shrub–tundra regions; or the quantification of beaver dams and lodges.