Geometry of last glacial sorted nets from high-resolution airborne data.

Relict periglacial sorted nets are valuable indicators of past permafrost and climate states, but their field studies are usually challenging due to high time requirements and poor visibility. This study therefore tests the suitability of high-resolution digital elevation model and aerial photographs with a horizontal resolution of 0.5 m and 0.2 m, respectively, to map and analyse the geometry of 2000 sorted nets from the Last Glacial Period at one site in the Krkonoše Mts., Czech Republic. Since the sorted nets occupy an area of ∼1.79 ha, the density of their network accounts for ∼1117 cells per hectare. The sorted nets have a mean diameter of 3.46 ± 0.83 m, a mean height of 0.28 ± 0.10 m, and a mean estimated sorting depth of 0.94 ± 0.22 m. The number of sides ranges between three and ten, but averages 5.99 ± 1.13, and 83.4 % of the sorted nets are pentagonal to heptagonal, and their sides mostly meet at three-way junctions at a mean angle of 119.9 ± 25.6°. However, isometric sorted nets are rather rare as a length-to-width ratio attains 1.47 ± 0.28. The estimated sorting depth indicates that permafrost superimposed by ∼1 m thick active layer occurred at the study site at the end of the Last Glacial Period. Generally, the remotely sensed parameters of the sorted nets are consistent with field data previously collected at the same study site and literature reports including those on past regional environmental conditions. This proves the utility of the high-resolution airborne data to map and analyse the geometry of large sets of sorted patterns, which would be difficult in conventional field surveys. Consequently, remote sensing could bring a wealth of new information on sorted patterned ground and its characteristics from past and present periglacial landscapes, and aid in past permafrost and climate modelling. The collected dataset of the parameters of the sorted nets could also have many other applications such as for choosing an effective sample size or the minimum network size to be included in statistical comparisons in future patterned-ground surveys, and for validating automated mapping and/or delineation tools and models of patterned-ground growth. • Last glacial sorted nets were examined using high-resolution (<0.5 m) airborne data. • Length, width, and diameter of sorted nets differ by <13 % from ground-truth data. • Height of sorted nets is ∼35 % lower compared to ground-truth data. • Sorted nets developed within ∼1 m thick active layer over permafrost. • At least ∼30–100 sorted nets need to be sampled for their parameter errors <5 %. [ABSTRACT FROM AUTHOR]