Development of Data Assimilation Methods for Mesoscale Processes and General Circulation in the Ocean

1995-96由于数据同化是揭示海洋学过程的一种新方法，因此主要调查员和共同调查员之间交流信息极为有用。因此，我们于7月在筑波举行了一次会议，邀请了其他研究人员，共同审查了现有的方法和搜索数据集。PI访问了加拿大，并对俄罗斯CI（居住在加拿大）开发的伴随方程方法进行了研究。他们访问日本一个半月，参加PI的项目。他们参观了日本的许多研究机构，并与日本科学家交流了知识。PI和俄罗斯CI向海洋学杂志提交了一篇论文。一名CI（Kamachi）访问了挪威，调查卡尔曼滤波方法。这次访问是非常必要的，因为在日本没有这方面的专家。1996 - 97继第一年的第一次会议之后，我们举行了第二次会议， ...更多信息 在陆津举行的国际会议上，邀请了来自不同国家的五位CI。参加者总数超过40人。与会者就比第一年更先进的数据同化方法交流了信息。一些建议包括（1）在伴随方法中，具有更简单物理的动力学模型可能更适合作为约束模型，以及（2）在卡尔曼滤波方法中，在处理一个强非线性系统时，有必要消除不稳定性。PI访问了夏威夷大学，交流北太平洋大气环流资料同化的信息。可以使用伴随方法从卫星高度计数据重建地下场，（2）可以从表面速度数据重建陆架环流，（3）包括高度计数据和漂移器数据两者以重建大尺度环流是有用的，（四）通过同化表层资料，可以重建表层混合层下的海洋场，已经实现了。信息交流对于PI和CI在初始阶段开展研究非常有用。少

1995-96Since data assimilation is a new approach for revealing oceanographic processes, it was extremely useful for the principal investigator (PI) and the co-investigators (CI) to exchange the information that they had among them. Hence, we held a meeting in Tsukuba in July, in which the other researchers were also invited and worked together to review the existing methods and search data sets.The PI visited Canada and did Research on the adjoint equation method taht Russian CI (living in Canada) developed. They visited Japan for one and half months to participate in the PI's projects. They visited many research institutes in Japan and exchanged their knowledge with Japanese scientists. The PI and Russian CI submitted a paper to Journal of Oceanography. One CI (Kamachi) visited Norway to investigate the Kalman filter method. This trip was highly necessary, because there was no specialist in Japan in this field.1996-97Following the first meeting in the first year, we held the second, m … More ore international meeting in Mutsu by inviting five CI from various countries. The total number of the participants wa more than 40. The participants exchanged the information on the more advanced methods of data assimilation than the first year. Some suggestions include (1) in the adjoint method, a dynamical model with a simpler physics may be more suitable as a constraint model, and (2) in the Kalman filter method, it is necessary to remove instability when a strongly nonlinear system is dealt with.The PI visited University of Hawaii to exchange information on data assimilation for general circulation in the North Pacific.Outcomes from this projects are(1) it is possible to reconstruct subsurface field from satellite altimeter data using the adjoint method,(2) it is possible to reconstruct shelf circulation from surface velocity data,(3) it is useful to include both altimeter data and drifter data to reconstruct large scale circulation,(4) it is possible to reconstruct the oceanic field beneath the surface mixed layr by assimilating surface data.The objectives initially listed in the proposal have been achieved. The information exchange was extremely useful for the PI and CI to develop their research at the initial stage. Less