Assessment of timberline boundary shift in the mountainous ecosystems of Subpolar Urals from multispectral satellite imagery over the last 40 years

Abstract. Shifts in the habitat boundaries of woody plant species represent one of the continental-scale consequences of climate change. Mapping these shifts and quantitatively evaluating them is essential for an appropriate assessment of the carbon balance. Here we present a stepwise procedure including the selection and processing of multispectral remote sensing data from Landsat 4–9 on the Google Earth Engine servers. We show explicitly how these data can be used to assess the timberline boundary shift in the mountainous ecosystems of the Subpolar Urals (exemplified for the Sablya ridge) using such methods as pseudo–invariant feature matching, robust regression, principal component analysis, and logistic regression-based classification, resulting in high classification accuracy indicated by Intersection over Union (IoU) above 0.9. The quantitative evaluation of the overall forest advancement area from 1960 to 2024 — based on observations collected between 1987 and 2024 — was estimated at 4.82 km², in a reasonable agreement with the expert delineation estimate of 5.6 km².