Integrating SfM, LiDAR and TLS for High‐Resolution Monitoring of Palsa Surface Changes and Degradation Dynamics

Understanding fine‐scale surface changes is crucial for quantifying ecosystem dynamics and detecting landscape changes. Palsas, comparably small peatland elevations with an ice‐rich permafrost core, undergo natural cycles of growth and degradation. They can also serve as indicators of climate change in subarctic regions. Because they typically cover an area of only a few hundred square metres at most, satellite remote sensing is insufficient for monitoring their detailed development. Unoccupied aerial systems (UAS) and ground‐based remote sensing provide the necessary spatial resolution, and combining multiple sensor types can further improve data quality. In this study, we monitored the annual surface changes of two degrading palsas in northern Finnish Lapland from 2022 to 2024. We captured UAS RGB‐based structure‐from‐motion (SfM), UAS LiDAR and terrestrial laser scanning (TLS) point clouds and validated the resulting digital terrain models (DTMs) using RTK‐GNSS reference measurements. Active layer thickness (ALT) was measured at each GNSS point to characterize thaw depth and to delineate an ALT‐constrained frozen extent. Based on the DTMs, we quantified changes in palsa area, mean height, and volume above a predefined palsa base. These DTM‐based metrics were evaluated for the entire palsa surface and, separately, for the part of the palsa where ALT measurements indicated the presence of a remaining frozen core. The accuracy assessment shows clear advantages of merging SfM point clouds with LiDAR‐ or TLS‐derived datasets. The highest accuracy was achieved at the smaller and flatter palsa, suggesting that high point cloud density and low surface complexity increase accuracy. UAS LiDAR showed a consistent positive elevation bias, likely reflecting a combination of slope‐ and footprint‐related effects and limited penetration of moist moss and peat surfaces. Both palsas lost volume, area and average height during the study period. The larger palsa decreased in total volume by 20% across its entire extent and by 23% within the ALT‐constrained frozen extent; the smaller palsa lost 13% and 18%, respectively. These results underscore the acceleration of palsa degradation under current climatic conditions and demonstrate that multisensor approaches significantly improve high‐resolution monitoring of palsa surface change in permafrost landscapes.