Collaborative Research: Exploring centennial changes in ocean circulation with SODA

该项目扩展了正在进行的工作（SODA：简单海洋数据同化），探索利用平行海洋计划（POP）重建百年海洋环流分析的建模和数据同化的潜力和局限性。研究领域是全球海洋，重点是观测记录更为丰富的上层海洋。主要的时间段是上个世纪下半叶（1950-2007），如果能够确定在上个世纪上半叶有足够的观测来提供对海洋状况的可靠估计，则着眼于将其扩展到上个世纪。第一项工作是收集再分析所需的观测数据和地表强迫数据集。事实上，这个项目的动机是气象学家为进行百年大气再分析所做的相应努力。将对用于同化的信息的误差统计进行彻底的调查。第二项工作是探索一种基于局部集成变换卡尔曼滤波器的更复杂的同化方案的变化。这种变化将允许数据同化的属性，如温度-盐度误差协方差在整个再分析期间演变，而不是先验地固定。还将注意改进混合层结构的表示。再分析实验的结果将用于探索各种海洋气候信号的结构和演变，如厄尔尼诺现象？太平洋的南方涛动（ENSO）和全球海洋的年代际变化。总体目标是探索海洋数据同化在多大程度上可用于为海洋产生参考气候，该参考气候反过来可作为检验其他物理、化学和生物分析产品的基础，为观测集提供大尺度背景，以及为海洋研究提供参考。S对气候变率的贡献。知识价值：这项研究将使人们了解对海洋物理状态进行百年一次的重新分析的潜力，重点是1000米以上。将编制百年再分析，并将其与强迫和海洋数据集一起提供给社区。智力上的价值在于，这种重新分析有可能对海洋提出新的问题。S在气候变率中的作用。这些结果将用于对季节、年际和年代际时间尺度上的变率进行诊断性检查，包括对大气气候变率协变的检查。最后，将海洋再分析结果与独立观测集进行比较，以了解再分析中的不确定性。更广泛的影响：SODA再分析目前正在被世界各地的研究人员使用，由于覆盖时间更长，并且由于新的同化方案，应该会对这些新的再分析产生相当大的兴趣。这项工作产生的一组再分析实验将提供给社区，并将为学生和研究人员提供许多探索海洋气候变化的机会。除了对马里兰大学和德克萨斯农工大学的研究生进行海洋数据同化方面的培训外，再分析还可以作为其他机构研究生的基础，并作为课堂教学的资源。

This project extends an on-going effort (SODA: Simple Ocean Data Assimilation) to explore the potential as well as the limitations of modeling and data assimilation directed toward reconstruction of a centennial analysis of ocean circulation using the Parallel Ocean program (POP). The domain is the global ocean, with an emphasis on the upper ocean where the observational record is more substantial. The main time period is the second half of the last century (1950-2007) with an eye to expand it to the last century if it can established that there are enough observations in the first half of the century to provide reliable estimate of the state of the ocean.. The first effort is to assemble the observations and surface forcing data sets needed for this reanalysis. Indeed, the project is motivated by a corresponding effort by meteorologists to produce a centennial atmospheric reanalysis. A thorough investigation of the error statistics of information that goes into the assimilation will be conducted. The second effort is to explore a change to a more sophisticated assimilation scheme based on the Local Ensemble Transform Kalman Filter. This change would allow properties of the data assimilation such as temperature-salinity error covariances to evolve throughout the reanalysis period, rather than being fixed a priori. Attention will also be focused on improving representation of the structure of the mixed layer. The results of the reanalysis experiments will then be used to explore the structure and evolution of a variety of ocean climate signals such as the El Nino?Southern Oscillation (ENSO) in the Pacific Ocean and decadal variability in the global oceans. The overall goal is to explore the extent to which ocean data assimilation may be used to produce a reference climate for the oceans that in turn can serve as the basis for examining other physical, chemical, and biological analysis products, for providing the large-scale context for observation sets, as well as for studies of the ocean?s contribution to climate variability.Intellectual merit: This research will result in understanding of the potential of a centennial reanalysis of the physical state of the ocean with emphasis on the upper 1000m. Centennial reanalyses will be produced and made available to the community, along with the forcing and ocean data sets. The intellectual merit lies in the potential of this kind of reanalysis to open up new questions regarding the ocean?s role in climate variability. The results will be used to carry out diagnostic examination of variability on seasonal, interannual, and decadal timescales including examination of covariations of atmospheric climate variability. Finally, the ocean reanalysis product will be compared to independent observation sets in order to learn about the uncertainties in the reanalysis.Broader impacts: SODA reanalyses are currently being used by researchers from around the world and there should be considerable interest in these new reanalyses, both because of the longer period of coverage and because of the new assimilation scheme. The set of reanalysis experiments resulting from this work will be made available to the community and will offer many opportunities for students and researchers to explore ocean climate variability. In addition to training graduate students at the University of Maryland and at Texas A&M University in ocean data assimilation, the reanalysis can be used as the foundation for graduate students at other institutions and as a resource for classroom teaching.