Impact of global warming rate on permafrost degradation

The IAP RAS climate model of intermediate complexity is used to analyze the sensitivity of the area of continuous potential permafrost Scont to the rate of global temperature variation Tgl in experiments with greenhouse-gas increases in the atmosphere. The influence of the internal variability of the model on the results is reduced by conducting ensemble runs with different initial conditions and analysis of the ensemble means. Idealized experiments with a linear or exponential dependence of the concentration of carbon dioxide in the atmosphere have revealed an increase in the magnitude of the temperature-sensitivity parameter of the area of continuous potential permafrost, kcont (= Scont, 0t-1dScont/dTgl, where Scont, 0 is the present value of Scont). With a decrease in the linear trend coefficient of Tgl from about 3 to about 2 K/100 yr, this parameter varies from approximately −0.2 to −0.4 K−1. With an even slower change in global temperature, kcont virtually does not vary and remains close to the value obtained from paleoreconstructions of the past warm epochs. Such a dependence of kcont on the rate of global warming is related mainly to the fact that the more rapid increase in Tgl leads to a slower response over high-latitude land. The contribution from changes in the annual temperature cycle, though comparable in the order of magnitude, is about one-third as large as the contribution from the variation of the latitudinal structure of the response of annual mean temperature. The total reduction in the annual cycle of temperature during warming partly compensates for the effect of the annual mean temperature rise, thus decreasing the magnitude of kcont. In numerical experiments with greenhouse gas changes in accordance with SRES scenarios A2 and B2 and scenario IS92a, there is also a monotonic increase in the magnitude of the normalized parameter of temperature sensitivity of the area of continuous permafrost with a decrease in the growth rate of global temperature. For scenarios A2-CO2, IS92a-GHG, IS92a-CO2, B2-GHG, and B2-CO2, its value is almost indistinguishable from the steady-state asymptotic value of −0.4 K−1. For A2-GHG, the magnitude of kcont turns out to be far less (kcont ≈ −0.3 K−1).