Collaborative Research for WOCE-AIMS: Global Velocity Estimation Project

欧文斯/ mcwilliams这个合作项目将尝试利用在世界海洋环流实验（WOCE）现场操作阶段之前和期间获得的水下浮子、海流仪和海面漂浮器的速度数据、水文剖面和海面高度测量数据来估计北大西洋的大尺度三维速度场。随着GRACE任务的重力测量、JASON任务的测高、ARGO漂浮计划的速度和T和S剖面等新测量数据的可用，它们将被纳入分析。速度及其预期误差的估计将基于预期测量误差和时空协方差和/或基函数的先验说明，使用广义最小二乘误差最小化来进行。将使用先验统计的层次结构。它们将根据拉伸盆地基函数的一般光谱形状和为其他目的运行的数值一般环流模式的模拟分析以及测量结果的统计分析来确定。模型协方差和基函数最初将精确地加强地转动力学，并在必要时添加额外的地转过程，例如在地表边界层和赤道附近。还将检查直接从模拟中计算的稍微不受约束的协方差的使用。将估计的后验统计量与先验统计量进行比较，以评估估计过程和模型的有效性。将作出一系列综合程度不断扩大的估计。最初，这些估计将从高采样区域的平均环流中获得，即大西洋气候和环流实验，然后扩展到第一个盆地，然后是全球地图。当我们对这一程序有信心时，我们也将制作速度随时间变化的地图。最终目标是创建一个活的评估库和软件系统，能够用户选择变量并使用新数据进行重新评估。这个项目是对WOCE的贡献

0137122/0137077 Owens/McWilliamsThis collaborative project will try to estimate the large scale, three-dimensional velocity field for the North Atlantic using velocity data from sub-surface floats, current meters, and surface drifters, hydrographic profiles, and sea surface height measurements taken prior to and during the field operations phase of the World Ocean Circulation Experiment (WOCE). As new measurements, such as the gravimetry from the GRACE mission, altimetry from the JASON mission, and velocities and T and S profiles from the ARGO float program become available, they will be incorporated into the analysis. The estimates of the velocities and their expected errors will be made using generalized least-square error minimization based on a priori specification of expected measurement errors and space-time covariances and/or basis functions. A hierarchy of a priori statistics will be used. They will be determined from general spectral shapes for stretched basin basis functions and analyses of simulations from numerical general circulation models run for other purposes, as well as statistical analyses from the measurements. The model covariances and basis functions initially will exactly enforce geostrophic dynamics, with additional ageostrophic processes added where necessary, such as in the surface boundary layer and near the Equator. The use of slightly less constrained covariances calculated directly from the simulations will also be examined. The a posteriori statistics of the estimates will be compared with the a priori ones to evaluate the estimation procedure and the validity of the models. A sequence of estimations with expanding comprehensiveness will be made. Initially these estimates will be obtained for the mean circulation in highly sampled regions, i.e., the Atlantic Climate and Circulation Experiment and then expanded to first basin then global maps. As we gain confidence in the procedure, maps for temporally varying velocities will also be made. The ultimate goal is to create a living estimation library and software system, capable of user-chosen variants and re-estimation with new data. This project is a contribution to WOCE