Permafrost Zones and Permafrost Temperatures

The concept of permafrost zones has predominated in mapping permafrost at national scale since the study of permafrost began in Russia in the 1880s (Nikiforoff 1928, Heginbottom 1984), and most permafrost maps delineate zones on the basis of spatial continuity. 

Continuous permafrost is defined as:

"Permafrost occurring everywhere beneath the exposed land surface throughout a geographic region with the exception of widely scattered sites, such as newly deposited unconsolidated sediments, where the climate has just begun to influence the ground thermal regime causing the formation of continuous permafrost." (NRCC 1988)

Discontinuous permafrost is defined as

"Permafrost occurring in some areas beneath the exposed land surface throughout a geographic region where other areas are free of permafrost." (NRCC 1988)

More recent maps (Harris 1986, Allard and Seguin 1987, Heginbottom et al. 1995) have attempted to define permafrost zones explicitly in terms of spatial extent. For example, in the map used for comparison in this report (Heginbotton et al. 1995) the following distinctions are used:

Permafrost Zone

Areal Extent

Continuous Permafrost

> 90%

Extensive Discontinuous


Sporadic Discontinuous


Isolated Patches


Temperature Boundaries

Because of the scarcity of field observations, sufficient to define the spatial continuity (extent) of permafrost in an area, air and ground temperatures have long been used as surrogates to delineate permafrost zones on national scale maps. For example, Nikiforoffs (1928) map is based solely on the -2oC mean annual air isotherm. Brown (1967) used climatic data for the region east of James Bay, using air temperature relationships developed in Siberia. Brown (1970) originally delineated continuous permafrost based on the Russian convention that the regional mean annual ground temperature is 5C (Brown, 1970). In the absence of sufficient ground temperature data, isotherms of air temperature have been used to delineate permafrost zones.

Discussions over the years of the relation between air temperature and permafrost temperature (e.g. Brown 1960, Williams and Smith 1989) have recognized that regional climate parameters (such as screen air temperature or freezing/thawing indices) have an imprecise relationship to local permafrost temperatures. Brown (1970) observed that the permafrost temperature can be as much as several degrees warmer than MAAT. On this basis, he stated that the southern limit of permafrost corresponds roughly with the 1oC mean annual air isotherm. He placed the continuous boundary on the -8oC isotherm but noted that the band of continuous permafrost along the coast of Hudson Bay west of James Bay appeared to be anomalous in this regard. French and Slaymaker (1993) state that southern limit of continuous permafrost corresponds with a MAAT of 6 to 8C, with the southern discontinuous limit at about 1C.

Correspondence between ground temperature and permafrost continuity

Using the maps of permafrost zones and of regional ground temperature in Heginbottom et al. (1995), the relation between permafrost continuity and ground temperatures can be examined. The two maps were compared by subdividing both into 10 x 10 km grid squares and cross-tabulating the ground temperature zones from one map with permafrost zones of the other. The following table shows the resulting breakdown of permafrost zones by ground temperature categories:



-15 to 10

-10 to -5

-5 to -2

-2 to 0



Continuous (90-100%)








Extensive Discontinuous (50-90%)








Sporadic Discontinuous (10-50%)








Isolated Patches (0-10%)








No permafrost








Mountain Permafrost
















Note: These percentages are for the areal extent of regions, and not the areal extent of permafrost per se. For example, while the "isolated patches" region covers 11.4% of Canada, permafrost within this region actually covers perhaps 0.5% of the land area.

The temperature classes for each of the permafrost zones correspond to the values expected from the classic permafrost literature. The two major subdivisions of the discontinuous permafrost zone are divided at the 0oC isotherm. However, significant proportions of the continuous permafrost zone are warmer than the Russian mapping criterion of -5oC mentioned above. The results suggest that 34% of the continuous permafrost zone has ground temperatures warmer than 5oC. Comparison of the zone boundaries on the maps indicates that the anomalously warm continuous permafrost is mostly in four regions:

Of these, the Keewatin anomaly may be an artifact of the mapping method, since the boundary in this region is based on ground temperature at a single borehole.