Geometry of LGM polygonal sorted patterns analysed using high-resolution airborne data (Krkonoše Mountains, Czech Republic)

Relict sorted patterns are valuable indicators of past permafrost and climate evolution, but their detailed terrain explorations are usually challenging due to high time requirements and poor pattern visibility. Here, we test the applicability of high-resolution airborne data to map and analyse the geometry of LGM polygonal sorted patterns at one site in the Krkonoše Mts., Czech Republic. We delineated a total of 2000 sorted patterns using colour contrasts between their elevated centres and bordering troughs discernible on a LiDAR digital elevation model with a resolution of 0.5 m and on true-colour orthogonal aerial photographs with a resolution of 0.2 m. Since the patterns occupy an area of ~1.96 ha, the density of their network accounts for ~1019 cells per hectare. The patterns have a diameter of 3.59±0.95 m, a height of 0.30±0.11 m, and an estimated sorting depth of 1.00±0.26 m. The number of pattern sides ranges between three and ten, but 82 % of the patterns are pentagonal to heptagonal, and their sides mostly meet at three- or four-way intersections at an angle of 120±24°. However, isometric patterns are rather rare as a length-to-width ratio attains 1.48±0.30. Generally, the remotely-sensed pattern attributes are consistent with ground-truth data previously collected at the study site, which proves the utility of high-resolution airborne data to rapidly map and complexly analyse the geometry of large sets of relict landforms over extensive areas that could not be done by conventional terrain surveys. The sorting depth indicates that permafrost superimposed by ~1 m thick active layer occurred at the study site during the LGM, which can be further used for past permafrost and climate modelling. The dataset can also have many other applications such as for validating automated pattern mapping/delineation tools and pattern growth models or for choosing an effective sample size for future surveys.

The research is financially supported by the Czech Science Foundation, project number 21-23196S.