多年冻土退化对生态系统、水资源、人类基础设施和气候等具有重要影响。厘清多年冻土退化的原因是应对这些影响的前提。利用CMIP5多模式数据，并结合地表冻结指数模型，我们首次研究了人类和自然外强迫对20世纪北半球近地表多年冻土变化的影响。在人类温室气体强迫下，近地表多年冻土面积显著减少，趋势为0.46×106 km2/10年，与观测和全强迫结果一致。在人类气溶胶强迫下，近地表多年冻土面积增加，趋势为0.25×106 km2/10年。在自然强迫下，近地表多年冻土趋势较弱，但具有明显的年代际变化。全强迫和温室气体强迫信号能在观测的近地表多年冻土变化中检测得到，但自然和气溶胶强迫信号不能被检测，表明观测的近地表多年冻土退化归因于温室气体强迫作用。
Given the current confirmed permafrost degradation and its considerable impacts on ecosystems, water resources, infrastructure and climate, there is great interest in understanding the causes of permafrost degradation. Using the surface frost index (SFI) model and multimodel data from the fifth phase of the Coupled Model Intercomparison Project (CMIP5), this study, for the first time, investigates external anthropogenic and natural forcing impacts on historical (1921–2005) near-surface permafrost change in the Northern Hemisphere. The results show that anthropogenic greenhouse gas (GHG) forcing produces a significant decrease in the area of near-surface permafrost distribution at a rate of 0.46 × 106 km2 decade−1, similar to observations and the historical simulation (ALL). Anthropogenic aerosol (AA) forcing yields an increase in near-surface permafrost distribution area at a rate of 0.25 × 106 km2 decade−1. Under natural (NAT) forcing, there is a weak trend and distinct decadal variability in near-surface permafrost area. The effects of ALL and GHG forcings are detectable in the observed change in historical near-surface permafrost area, but the effects of NAT and AA forcings are not detected using the optimal fingerprint methods. This indicates that the observed near-surface permafrost degradation can be largely attributed to GHG-induced warming, which has decreased the near-surface permafrost area in the Northern Hemisphere by approximately 0. 21 × 106 km2 decade−1 on average over the study period, according to the attribution analysis.
图 1 近地表多年冻土面积的变化。（a）全强迫；（b）温室气体；（c）自然强迫；（d）气溶胶强迫。阴影代表一个标准差。
Figure 1. Temporal changes in near-surface permafrost area under (a) historical simulation (ALL), (b) greenhouse gas (GHG), (c) natural (NAT), and (d) anthropogenic aerosol (AA) experiments. The simulated near-surface permafrost area has been diagnosed by the SFI model driven by CRU data climatology plus anomalies of each model. Shaded areas represent one standard deviation across models. The black line in panel (a) represents the observed change in near-surface permafrost area diagnosed by the SFI model driven by CRU data. Linear trends (Trend) (106 km2 decade−1) in ensemble-mean near-surface permafrost area for 1921–2005 from observations and four external forcings are given in the bottom of each panel.
Guo Donglin, Sun Jianqi, Huixin Li, Tingjun Zhang, Vladimir E. Romanovsky, 2020: Attribution of historical near-surface permafrost degradation to anthropogenic greenhouse gas warming. Environmental Research Letters, 15, 084040.