A Comprehensive Database of Thawing Permafrost Locations Across Alaska
Bibliographic record
Abstract
Abstract. The Arctic is warming nearly four times faster than the global average, leading to widespread permafrost thaw degradation with profound implications for ecosystems, infrastructure, and global climate feedbacks. While gradual permafrost thaw occurs over decades, abrupt thaw events – such as thermokarst formation or retrogressive thaw slumps – can rapidly alter ecosystems and severely damage infrastructure. Although abrupt thaw is increasingly widespread, comprehensive datasets that map its spatial distribution at regional scales for land managers and local governments are still lacking. To address this gap, we created the Alaska Permafrost Thaw Database, an open-access, collaborative database which compiles 19,540 permafrost thaw and thermokarst locations across Alaska from 44 sources, integrating field observations, remote sensing products, and the published literature. This database spans observations from 1950 through present and incorporates datasets of varying spatial resolution, ranging from field-based point measurements to remotely sensed products (1–125 m), providing statewide coverage across Alaska. The dataset includes abrupt thaw features and sites experiencing gradual top-down thaw that can help to support comparative analysis and predictive modeling. We used this database to explore relationships between thaw type (abrupt vs. non-abrupt) and topographic metrics (i.e., slope, relative elevation, and potential incoming solar radiation), analyze the distribution of various thaw features across Alaska’s major ecoregions, and compare the database to current spatial datasets of ground ice and Yedoma. Our analysis shows abrupt thaw features are more prevalent in lowlands and depressions while gradual top-down and lateral thaw features are more commonly associated with areas receiving higher potential incoming solar radiation such as south facing slopes and open clearings. We also found substantial mismatches between ice-driven thaw processes and existing ground ice and Yedoma maps, likely reflecting the coarse resolution of current mapping products relative to the fine-scale nature of field measurements and highlighting the limitations of current datasets for local-scale prediction. The database provides direct, empirical evidence of actively thawing and stable permafrost locations and can be used to inform and validate ground ice mapping. By comparing the database with physiographic characteristics and remotely sensed measurements, the database can guide future field campaigns in areas with little to no observations. As permafrost thaw transforms Arctic landscapes, high-resolution, accessible spatial data – such as our thaw database – will be critical for informing climate mitigation and adaptation strategies. The Alaska Permafrost Thaw Database is openly available at Zenodo (https://doi.org/10.5281/zenodo.16996415), which provides a link to the GitHub repository and access to all versions; this paper describes version 2.0.0.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.006 | 0.008 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".