【摘要】 准确估算陆地生态系统呼吸（Ecosystem respiration,RE）对全球陆地生态系统碳收支研究具有重要意义.模型模拟是估算陆地RE变化的一种常用手段.然而目前陆地生态系统过程模型的RE模拟尚未得到充分验证.基于耦合模式比较计划第五阶段（Coupled Model Intercomparison Project Phase 5,CMIP5）的通用陆面模型（Community land model,CLM）RE模拟结果和全球通量网（FLUXNET）66个站点的涡度相关通量观测数据（277条站点年数据）评估CLM模型对RE的模拟效果.结果表明:¹在空间尺度上,CLM低估了高纬度站点RE,高估了低纬度站点RE,但高纬度低估量更大导致空间格局整体低估（相对误差为-3.56%）.（2）在时间尺度上,CLM模型基本捕捉了RE的年际和季节变化,相关系数分别为0.60（P<0.001）和0.63（P<0.001）;CLM低估年尺度和月尺度的RE（以C计_）,绝对误差分别是-182.21 g m-² a-¹、-120.16 g m-² mon-¹,相对误差分别是-17.84%、-10.60%.（3）CLM模型对不同植被功能型的RE模拟效果不同,由优及差依次为混交林、常绿针叶林、草地、农田、落叶阔叶林、常绿阔叶林.本研究在时空尺度上量化了CLM模型的生态系统呼吸模拟误差,并分析了土壤呼吸Q10和MRbase参数以及土壤碳库模拟等因素的影响,可为CLM模型的生态系统呼吸模块参数优化提供依据,进而提升其模拟精度.（图4表3参80附图2附表2）更多还原

【Abstract】 Accurate estimation of terrestrial ecosystem respiration（RE）is of great significance to the study of the global terrestrial ecosystem’s carbon budget.Model simulation is a common method to simulate terrestrial RE changes.However,the RE simulation of the current terrestrial ecosystem’s process models has not been fully verified yet.In this study,we evaluated the RE simulated by the Community Land Model（CLM）using eddy covariance flux observations of 66 stations from FLUXNET（277 site-years）.The results showed that:（1）CLM underestimated RE at high latitude sites while overestimated it at low latitude stations.The magnitude of the former was larger than that of the latter,thus leading to the overall underestimation of RE（the relative error was-3.56%）.（2）At the temporal scale,CLM roughly captured the interannual and seasonal variationof RE.The correlation coefficients were 0.60（P<0.001）and 0.63（P<0.001）,respectively.CLM underestimated the annual and monthly RE,with the absolute error of 182.21 g C m-² a-¹ and 120.16 g C m-² mon-1,respectively.The relative errors were17.84%and 10.60%,respectively.（3）The simulation results of the different plant functional types from the best to worst were ranked as mixed forest,evergreen coniferous forest,grassland,farmland,deciduous broadleaved forest,and evergreen broadleaved forest.This study quantified the error of CLM-RE simulation at the spatial-temporal scale and analyzed its influence factors,such as soil respiration Q10,MRbase parameters,and soil carbon reservoir simulation.The results can help in optimizing CLM-RE and improving the accuracy of RE simulations.更多还原