Data from 'Local Regions Associated With Interdecadal Global Temperature Variability in the Last Millennium Reanalysis and CMIP5 Models'

<strong>Abstract from '<em>Local Regions Associated With Interdecadal Global Temperature Variability in the Last Millennium Reanalysis and CMIP5 Models</em>':</strong> Despite the importance of interdecadal climate variability, we have a limited understanding of which geographic regions are associated with global temperature variability at these timescales. The instrumental record tends to be too short to develop sample statistics to study interdecadal climate variability, and Coupled Model Intercomparison Project, Phase 5 (CMIP5) climate models tend to disagree about which locations most strongly influence global mean interdecadal temperature variability. Here we use a new paleoclimate data assimilation product, the Last Millennium Reanalysis (LMR), to examine where local variability is associated with global mean temperature variability at interdecadal timescales. The LMR framework uses an ensemble Kalman filter data assimilation approach to combine the latest paleoclimate data and state-of-the-art model data to generate annually resolved field reconstructions of surface temperature, which allow us to explore the timing and dynamics of preinstrumental climate variability in new ways. The LMR consistently shows that the middle- to high-latitude north Pacific and the high-latitude North Atlantic tend to lead global temperature variability on interdecadal timescales. These findings have important implications for understanding the dynamics of low-frequency climate variability in the preindustrial era.《近千年再分析与CMIP5模式中与年代际全球温度变率相关的局地区域》论文摘要：尽管年代际气候变率具有重要研究价值，但目前我们对该时间尺度下与全球温度变率相关联的地理区域仍知之甚少。器测记录往往时长过短，难以构建样本统计量以研究年代际气候变率；而耦合模式比较计划第五阶段（Coupled Model Intercomparison Project Phase 5, CMIP5）的气候模式，对于哪些区域对全球平均年代际温度变率的影响最为显著这一问题，也存在显著分歧。本研究借助一款全新的古气候数据同化（paleoclimate data assimilation）产品——近千年再分析数据集（Last Millennium Reanalysis, LMR），探究年代际时间尺度下局地变率与全球平均温度变率相关联的区域。LMR框架采用集合卡尔曼滤波（ensemble Kalman filter）数据同化方法，融合最新古气候数据与当前最先进的模式数据，生成地表温度的逐年分辨率场重建结果，使我们能够以全新方式探究器测前气候变率的发生时序与动力学机制。LMR的结果一致表明，中高纬度北太平洋与高纬度北大西洋在年代际时间尺度上往往领先于全球温度变率。这些研究结果对于理解前工业时代低频气候变率的动力学机制具有重要意义。