About 25% of the land mass of the Northern Hemisphere is underlain by permafrost, including large regions of Canada, China, Russia and Alaska, with smaller permafrost areas in mountain chains of many other countries in both the Northern and Southern Hemisphere (Brown et al., 1997; Zhang et al., 1999). Permafrost in large part depends on climate. Over half of the world's permafrost is at temperatures a few degrees below 0°C. Temperature variations in near-surface permafrost (20 to 200 m depth) can be used as a sensitive indicator of the inter-annual and decade-to-century climatic variability and long-term changes in the surface energy balance (Lachenbruch and Marshall, 1986; Lachenbruch et al., 1988; Clow et al., 1991; Beltrami and Taylor, 1994; Majorowicz and Judge, 1994). Very small changes in surface climate can produce important changes in permafrost temperatures. Lachenbruch and Marshall (1986) used climate reconstructions from deep (>125 m depth) temperature measurements in permafrost to show that there has been a general warming of the permafrost in the Alaskan Arctic of 2 to 4°C over the last century.
Evidence of change in the southern extent of the discontinuous permafrost zone in the last century has also been recorded. In North America, the southern boundary of the discontinuous permafrost zone has migrated northward in response to warming after the Little Ice Age, and continues to do so today (Thie, 1974; Vitt et al., 1994; Halsey et al., 1995; Laberge and Payette, 1995; French and Egorov, 1998). In China both an increase in the lower altitudinal limit of mountain permafrost and a decrease in areal extent have been observed (Wang et al., 2000).
Long-term monitoring of shallow permafrost began in earnest in the last few decades. Recent analyses indicate that permafrost in many regions of the earth is currently warming (Gravis et al., 1988; Haeberli et al., 1993; Osterkamp, 1994; Pavlov, 1994; Wang and French, 1994; Ding, 1998; Sharkhuu, 1998; Vonder Mühll et al., 1998; Weller and Anderson, 1998; Osterkamp and Romanovsky, 1999; Romanovsky and Osterkamp, 1999). However, the onset, magnitude (from a few tenths to a few degrees centigrade) and rate of warming varies regionally, and not all sites in a given region show the same trend (Osterkamp and Romanovsky, 1999). This variability, as well as short-term (decadal or less) trends superimposed on long-term (century) trends, is briefly discussed in Serreze et al. (2000). There has also been evidence of recent permafrost cooling into the mid-1990s in parts of north-eastern and north-western Canada (Allard et al., 1995; Burn, 1998). However, there are regional data gaps, such as in the central and high Arctic in North America. A new international permafrost thermal monitoring network (Burgess et al., 2000) is being developed to help address these gaps and document the spatial and temporal variability across the globe.
Properties of the surface and the active layer (that having seasonal freezing and thawing) affect surface heat exchanges in permafrost regions. Other conditions remaining constant, the thickness of the active layer could be expected to increase in response to warming of the climate. A circumpolar network to monitor active-layer thickness at representative locations was developed in the 1990s to track long-term trends in active layer thickness (Nelson and Brown, 1997). Active layer thickness time-series are becoming available (Nelson et al., 1998; Nixon and Taylor, 1998), and evidence of increasing thaw depths is starting to be reported (Pavlov, 1998; Wolfe et al., 2000).
Other reports in this collection