First retrogressive thaw slump (RTS) inventory for the Kanin Peninsula (NW Russia)

We have chosen to manually interpret and digitize the inventory in order to produce it with the highest quality, robustness and representativeness (for both active and inactive features) at both local, regional, and pan-Arctic scale and context. The process was repeated twice to minimise bias in the mapping. Initially, it was conducted by a geomorphologist (the first author) and then quality checked and, if necessary, adjusted by an Arctic mapping Quaternary geologist (the second author) to ensure quality and make any necessary adjustments. For our study area, most RTS appear to be active or partly active landforms and can thus be considered recent. High-resolution (< 1 m) observations from ESRI Wayback Imagery, which archives all published versions of world imagery, were interpreted. The interpretation and digitising was conducted up to the maximum zoom level of 1:1000, with no mapping of features beyond this zoom level. This approach allows for unambiguous identification, in conjunction with the quality of the remote sensing data, while ensuring a homogeneous spatial accuracy of the final dataset. Our inventory contains 900 polygon features. They are categorized into 633 inactive and 267 active features based on morphology and vegetation patterns. The primary goal was not to create a complete RTS inventory but to develop a highly representative dataset for training machine learning models to automatically detect and classify RTS into active and inactive categories. Since the data have been produced and validated by two different scientists with different backgrounds, it can be said that they are of high quality and representativeness in local, regional and pan-Arctic contexts and scales. In detail, the inventory is comprised only of vector-polygons. There is one shapefile, named RTSinventory_Kanin_2022, which comprises all the features. Within the attribute table, there are 13 columns, which comprise details about each polygon/feature, as follows:

1 FID – ID showing the total number of polygons

2Shape – Polygon

3ID – each polygon has associated an ID

4Status – active/inactive

5Area – expressed in m².

6Sphericity – calculated as the ratio of the perimeter of the equivalent circle to the real perimeter of the RTS polygon

7Maximum Di (Maximum Distance) – calculated between two points along the polygon perimeter.

8Elongation (Index) – calculated as the maximum distance divided by the square root of the area.

9MeanAsp – the mean aspect where the respective RTS is located

1MeanElev – the mean elevation where the respective RTS is located

1MeanPlc – the mean planar curvature where the respective RTS is located

1MeanPrc – the mean profile curvature where the respective RTS is located

1MeanSlo – the mean slope where the respective RTS is located

SSome technical limitations must be mentioned due to the approach used to develop the inventory. The main limitation of the inventory is its temporal coverage. The polygons represent the reality on the ground as it was in 2022. However, this might be a strength for future change analysis. We did our best to overcome the potential negative effects on homogeneity of quality, taking into account the double validation process by two independent scientists. The limitation of ground truthing should be mentioned. There were no opportunities to be in the field to validate our inventory. However, based on previous experience on manual mapping of RTS in the pan-Arctic we consider the inventory of high quality. Another limitation is that RTS were not mapped beyond the 1:1000 zoom level.