Abstract:
Soil temperature is an important indicator of frozen ground status, driven at least partly by air temperature variability. In this study we apply singular spectrum analysis (SSA) to detect trends and oscillations in annual and seasonal time series of surface air temperature (SAT) and soil temperature (ST). We investigate soil temperatures at depths of 0.4, 1.6, and 3.2 m for five permafrost-occupied regions in Russia. We use SAT data for 1902-1995 and ST data for 1960-1990. The trends show an increase in annual SAT and ST from the end of the 1960s across all five regions, and this warming exceeds that of the preceding period in the Central Siberian Plateau and Transbaikalia. Oscillations in annual SAT and ST time series are coincident in the West Siberian Plain (7.7 year period) and in the western Central Siberian Plateau and Transbaikalia (2.7 year period). In general, on a seasonal basis, 2-3 year oscillations in ST and SAT are coincident during winter, spring, and autumn across the regions and are also evident in the annual ST time series in the Central Siberian Plateau and Transbaikalia. We also find a decadal oscillation (9.8 year periods), which is coincident for winter SAT and ST, over the western Central Siberian Plateau only. Although summer SAT and ST oscilations (5-8 year periods)(are coincident for all investigated territories (except to the east of the Lena River), in the annual ST time series they are identified only for the West Siberian Plain. We document the degree to which SAT controls ST in each region and explore the causative factors for some of the dominant periods. The maximum effect of SAT increases on permafrost may be observed in the Central Siberian Plateau and Transbaikalia, while elsewhere the observed ST increases do not threaten permafrost areas. Copyright 2006 by the American Geophysical Union.
