Modelling the spatial distribution of permafrost in Labrador–Ungava using the temperature at the top of permafrost

• Published in: Canadian Journal of Earth Sciences
• Main Authors: Way, Robert G., Lewkowicz, Antoni G.
• Other Authors: Gajewski, Konrad
• Format: Article in Journal/Newspaper
• Language: English
• Published: Canadian Science Publishing 2016
• Subjects: General Earth and Planetary Sciences; Canada; permafrost; Subarctic
• Online Access: http://dx.doi.org/10.1139/cjes-2016-0034; http://www.nrcresearchpress.com/doi/full-xml/10.1139/cjes-2016-0034; http://www.nrcresearchpress.com/doi/pdf/10.1139/cjes-2016-0034

Permafrost zonation in Labrador–Ungava ranges from very isolated patches through to continuous permafrost. Here we present a new estimate of the distribution of permafrost at high resolution (250 m × 250 m) using spatial numerical modelling supported by station data from 29 air and ground climate monitoring stations. Permafrost presence was estimated using a modified version of the temperature at the top of permafrost (TTOP) model. Mean ground surface temperatures were modelled using gridded air temperatures and a novel n-factor parameterization scheme that compensates for regional differences in continentality, snowfall, and land cover and is transferable to other Subarctic environments. The thermal offset was modelled using land cover and surficial material datasets. Predicted TTOP values for the average climate range regionally from −9 °C (for high elevations in northern Quebec) to +5 °C (for southeastern Labrador – Quebec). Modelling for specific temporal windows (1948–1962, 1982–1996, 2000–2014) suggests that permafrost area increased from the middle of the 20th century to a potential peak extent (36% of the total land area) in the 1990s. Subsequent warming is predicted to have caused a decrease in permafrost extent of one-quarter (95 000 km 2 ), even if air temperatures rise no further, providing air and ground temperatures equilibrate. Zonal boundaries derived by upscaling the high-resolution model are highly scale dependent, precluding direct comparison with the Permafrost Map of Canada that was generated without the use of geographic information system based analyses.