The dataset presents georeferenced historical images of CORONA (1964 and 1969) and HEXAGON (1972, 1977, 1982, 1984) satellites for 5 key sites in the West Siberian Arctic. Georeferenced images were used to analyze the dynamics of retrogressive thaw slumps in this region. CORONA image georeferencing was performed in the open-source software QGIS version 3.34.8, using the ESRI Satellite basemap as a reference layer. We used a 3rd polynomial transformation method. The number of ground control points (GCPs) varied depending on terrain characteristics, image properties, and the mean error (RMS, 2.7-3.1 m) reported by the QGIS georeferencing module, but was not less than 40 points for each image. As the resampling method, we used a nearest neighbor algorithm. CORONA pixelsize is 2.75m. HEXAGON imagery was also georeferenced in QGIS, following the same procedure as for CORONA imagery, with the ESRI Satellite basemap used as reference. We also used a 3rd polynomial transformation (1) with nearest neighbor resampling. However, in certain cases, we used the Thin Plate Spline method with cubic (4x4 kernel) resampling (2) instead. The choice of method depended on image-specific characteristics and was based on a visual assessment of the georeferencing quality: method (1) was suitable for most images, as they were of high quality and acquired in (near-)nadir orientation. For images acquired at an oblique angle or experiencing distortions, method (2) performed better. For each image, around 40-50 Ground Control Points (GCPs) were created. HEXAGON pixelsize is 1 m.

Funded by "Der Deutsche Akademische Austauschdienst (DAAD) - STIBET I", google.org " Impact Challenge on Climate Innovation to the Permafrost Discovery Gateway development", BMWK "ML4EARTH"