| Title | Permafrost thaw sensitivity prediction using surficial geology, topography, and remote-sensing imagery: a data-driven neural network approach |
| |
| Author | Oldenborger, G A ;
Short, N; LeBlanc, A -M |
| Source | Canadian Journal of Earth Sciences 2022 p. 1-17, https://doi.org/10.1139/cjes-2021-0117  Open Access |
| Image |  |
| Year | 2022 |
| Alt Series | Natural Resources Canada, Contribution Series 20210439 |
| Publisher | Canadian Science Publishing |
| Document | serial |
| Lang. | English |
| Media | paper; digital; on-line |
| File format | pdf |
| Province | Nunavut |
| NTS | 55K/16 |
| Lat/Long WENS | -92.5000 -92.0000 63.0000 62.7500 |
| Subjects | hydrogeology; general geology; permafrost; satellite imagery; topography; remote sensing; textural classifications; Rankin Inlet Group; permafrost thaw |
| Illustrations | location maps; geological sketch maps; digital elevation models; tables; photographs; distribution diagrams; graphs |
| Program | Climate Change
Geoscience Permafrost |
| Released | 2022 02 23 |
| Abstract | Seasonal or degradational thaw subsidence of permafrost terrain affects the landscape, hydrology, and sustainability of permafrost as an engineering substrate. We perform permafrost thaw sensitivity
prediction via supervised classification of a feature set consisting of geological, topographic, and multispectral variables over continuous permafrost near Rankin Inlet, Nunavut, Canada. We build a reference classification of thaw sensitivity using
process-based categorization of seasonal subsidence as measured from differential interferometric synthetic aperture radar whereby categories of thaw sensitivity are reflective of ground ice conditions. Classification is performed using a neural
network trained on both dispersed and parcel-based reference data. For Low, Medium, High, and Very High thaw sensitivity categories, generalized classification accuracy is 70.8% for 20.6 km2 of dispersed training data. In all cases, the majority
classes of Low and Medium thaw sensitivity are predicted with higher accuracy and more certainty, while the minority classes of High and Very High thaw sensitivity are underpredicted. Minority classes can be combined to improve accuracy at the
expense of a reduced level of discrimination. The two-class problem can be classified with an accuracy of 81.8%, thereby effectively distinguishing between stable and unstable ground. The method is applicable to similar Low-Arctic permafrost terrain
with geological and topographical controls on thaw sensitivity. However, generalized accuracy is reduced for parcel-based training, indicating that reference samples are not totally representative for inference beyond the parcel, and any deployment
of the network to other geographical regions would benefit from full or partial retraining with local data. |
| Summary | (Plain Language Summary, not published)
Thaw subsidence of permafrost affects the landscape, hydrology, and sustainability of permafrost as an engineering substrate. Classification of satellite
imagery provides a means of remotely predicting permafrost conditions. We develop and train a machine learning model to classify permafrost terrain into Low, Medium, High, and Very High thaw sensitivity categories. The model is trained by supervised
learning of geological, topographic, and multi-spectral variables in continuous permafrost near Rankin Inlet, Nunavut. The majority classes of Low and Medium thaw sensitivity are predicted with higher accuracy and more certainty, while the minority
classes of High and Very High thaw sensitivity are under-predicted. The two-class problem can be classified with an accuracy of greater than 81% dividing the terrain into thaw stable and unstable ground. |
| GEOSCAN ID | 329289 |
|
|