Assessment of the Vulnerability of Permafrost Carbon to Climate Change: A Sensitivity Analysis among Models

This activity is a comparison of how large-scale models represent permafrost carbon dynamics into the future (2010-2299). Model responses were evaluated at several temporal scales. To the extent possible, we standardized driver data and simulation procedures among the models. However, the protocol has been set up so that each model can build upon the procedures used to produce the outputs for historical analysis (1960- 2009) that was published in McGuire et al. 2016 (Global Biogeochemical Cycles 30:1015-1037, doi:10.1002/2016GB005405). Note that this comparison is an offline model comparison in which we assessed the sensitivity of the responses of the models to somewhat standardized forcing data. The activity compared among the models: Carbon dynamics: Predictions of average annual C fluxes (GPP, NPP, RH, CH4 fluxes, disturbance-related emissions, dissolved organic carbon export, lateral land used fluxes, etc.) and major pools for the northern permafrost region for the 2010-2299 period. Soil thermal dynamics: Predictions of annual soil thermal and hydrological dynamics at prescribed depths and the maximum annual active layer depth (in permafrost locations) for the 2010-2299 time period. The spatial simulation data for this project are are available through the National Snow and Ice Data Center (doi: 10.5067/ZRL5WJKN01XM).本研究旨在对比各类大尺度模型对未来（2010-2299年）多年冻土碳动态的表征情况。研究团队在多个时间尺度上对各模型的响应结果进行了评估。 我们尽可能对各模型的驱动数据与模拟流程进行了标准化处理。不过，本实验框架允许各模型基于此前用于生成1960-2009年历史分析输出的流程进行拓展——该历史分析成果已发表于McGuire等人2016年刊发于《全球生物地球化学循环（Global Biogeochemical Cycles）》的论文（第30卷：1015-1037，doi:10.1002/2016GB005405）。 需注意，本次对比属于离线模型对比，我们评估了各模型对一定程度标准化的强迫数据的响应敏感性。 本研究从以下维度对各模型进行对比： 1. 碳动态：针对2010-2299年时段，预测北半球多年冻土区的年均碳通量（包括总初级生产力（Gross Primary Productivity, GPP）、净初级生产力（Net Primary Productivity, NPP）、土壤异养呼吸（Rh）、甲烷（CH₄）通量、扰动相关排放、溶解性有机碳输出、侧向土地利用通量等）以及主要碳库。 2. 土壤热动态：针对2010-2299年时段，预测指定深度处的年度土壤热与水文动态，以及多年冻土区的年度最大活动层厚度。 本项目的空间模拟数据可通过国家冰雪数据中心（National Snow and Ice Data Center）获取（doi: 10.5067/ZRL5WJKN01XM）。