This is a Preprint and has not been peer reviewed. This is version 1 of this Preprint.
Downloads
Authors
Abstract
Retrogressive Thaw Slumps (RTS) and Active Layer Detachment Slides (ALD) are widespread thermal mass-wasting hillslope failures triggered by thawing permafrost. Despite increasing rates of these failures, knowledge about their pan-arctic spatial and temporal distribution remains limited. We present the Database of AI-detected Arctic RTS and ALD footprints (DARTS), the largest hillslope thermokarst database with over 43,000 individual features. which spans approximately 1.6 million km² for 2018 to 2023 and at least annual coverage in 2021 to 2023 for a ~900,000 km² region. DARTS is freely available in two processing levels: sub-annual and annually aggregated polygon footprints. The database was created with a fully automated workflow that leverages deep learning-based segmentation of PlanetScope multi-spectral imagery (3-5m spatial resolution) with minimum annual coverage. We validated DARTS using different datasets, achieving F1 scores ranging from 0 to 0.519, with more accurate results in RTS-rich areas. The DARTS database will be valuable for mapping, quantifying, and understanding hillslope thermokarst distribution and change over time across the circum-arctic permafrost region.
DOI
https://doi.org/10.31223/X5740Z
Subjects
Artificial Intelligence and Robotics, Computer Sciences, Earth Sciences, Geomorphology, Physical Sciences and Mathematics
Keywords
Permafrost, retrogressive thaw slumps, Dataset, pan-arctic, Deep learning, Artificial Intelligence, thermokarst, active layer detachment slides, hillslope thermokarst, remote sensing, Segmentation
Dates
Published: 2024-10-20 09:21
Last Updated: 2024-10-20 16:21
License
CC BY Attribution 4.0 International
Additional Metadata
Conflict of interest statement:
None
There are no comments or no comments have been made public for this article.