摘要：Ecosystems in China have been absorbing anthropogenic CO2 over the last three decades. Here, we assess future carbon uptake in China using models from phase 5 of Coupled Model Intercomparison Project under four socio-economic scenarios. The average of China's carbon sink from 2006 to 2100 represented by multimodel mean net ecosystem production(NEP) is projected to increase(relative to averaged NEP from 1976 to 2005) in the range of 0.137 and 0.891 Pg C a-1across differentscenarios. Increases in NEP are driven by increases in net primary production exceeding increases in heterotrophic respiration, and future carbon sink is mainly attributed to areas located in eastern China. However, there exists a considerable model spread in the magnitude of carbon sink and model spread tends to be larger when future climate change becomes more intense. The model spread may result from intermodel discrepancy in the magnitude of CO2 fertilization effect on photosynthesis, soil carbon turnover time, presence of carbon-nitrogen cycle and interpretation of land-use changes. For better quantifying future carbon cycle, a research priority toward improving model representation of these processes is recommended.

摘要：精准估算陆地生态系统碳汇对中国制定碳减排措施、实现碳中和目标有重要科学支撑作用.大气反演方法基于大气二氧化碳浓度观测数据,可有效估算地面二氧化碳通量的时空分布.目前中国陆地生态系统碳汇的大气反演估算仍有巨大不确定性,其中一个重要瓶颈在于缺乏足够的二氧化碳观测站点.本文研发了针对中国区域的大气反演框架,提出了二氧化碳观测网络的站点布设方案以最大程度降低中国陆地碳汇反演估算的不确定性.研究表明,在现有站点基础上,亟需在生长季植被生产力较高的东南、东北、华北和青藏高原地区增设二氧化碳观测站点.若在中国建设30个大气二氧化碳观测站点可将碳汇估算不确定性从每年10亿吨碳降低至3亿吨碳;建设60个站点可将不确定性进一步降低至每年2亿吨碳.这些站点将是“天-空-地”综合碳观测系统的重要组成部分,服务于中国二氧化碳收支反演和精准核算.