Climate model benchmarking with glacial and mid-Holocene climates
Climate model benchmarking with glacial and mid-Holocene climates
复制标题
DOI：
10.1007/s00382-013-1922-6
复制
发表时间：
2014-10
期刊：
影响因子：
4.6
通讯作者：
S. Harrison;S. Harrison;Patrick J. Bartlein;Simon C. Brewer;I. Prentice;I. Prentice;Meighan Boyd;I. Hessler;I. Hessler;K. Holmgren;K. Izumi;K. Willis
中科院分区：
文献类型：
--
作者：
S. Harrison;S. Harrison;Patrick J. Bartlein;Simon C. Brewer;I. Prentice;I. Prentice;Meighan Boyd;I. Hessler;I. Hessler;K. Holmgren;K. Izumi;K. Willis
过去的气候为模型预测气候变化的能力提供了一个测试。我们提出了一个全面的评估国家的最先进的模型对末次冰期最大和全新世中期的气候，使用重建的土地和海洋气候和模拟的古气候建模和耦合建模相互比较项目。尽管分辨率和复杂性更高，但较新的模型的性能并不比早期版本更好。气候敏感性的差异只能微弱地解释模式性能的差异。在冰川中，模型一直低估了陆地冷却（特别是在冬季），高估了海洋表面冷却（特别是在热带地区）。在全新世中期，模型通常低估了北方季风区的降水增加，高估了欧亚大陆中部的夏季变暖。模型通常捕捉大尺度的气候变化梯度，但再现空间模式的能力较为有限。尽管存在这些常见的偏差，但有些模型的表现要优于其他模型。
Past climates provide a test of models’ ability to predict climate change. We present a comprehensive evaluation of state-of-the-art models against Last Glacial Maximum and mid-Holocene climates, using reconstructions of land and ocean climates and simulations from the Palaeoclimate Modelling and Coupled Modelling Intercomparison Projects. Newer models do not perform better than earlier versions despite higher resolution and complexity. Differences in climate sensitivity only weakly account for differences in model performance. In the glacial, models consistently underestimate land cooling (especially in winter) and overestimate ocean surface cooling (especially in the tropics). In the mid-Holocene, models generally underestimate the precipitation increase in the northern monsoon regions, and overestimate summer warming in central Eurasia. Models generally capture large-scale gradients of climate change but have more limited ability to reproduce spatial patterns. Despite these common biases, some models perform better than others.