Collaborative Research: Stochastic Transport Models for the Coastal Ocean

PI寻求确定用于沿海海洋的最佳随机粒子输运（即拉格朗日）模型，该模型将提供高频（HF）雷达系统的丰富欧拉沿海环流数据与沿海资源管理人员所需的运输信息之间的必要联系，这些信息的任务是确定污染物、幼虫和海上丢失物体的命运。具体的研究目标是：(i)使用高频雷达和高分辨率漂浮物观察两个沿海地区的地表流场，（ii）开发精确的拉格朗日输运模型，以预测高频雷达场的轨迹，（iii）模拟轨迹并通过与原位漂浮物轨迹的比较来量化它们的技能，以及（iv）计算再分配核函数（RKFs），或连通性矩阵。并证明它们作为简单的概率近岸运输模型的效用。现有的轨迹模型将被修正，以考虑沿海海洋特有的流型。将开发一个具有大尺度平均分量(U)、代表潮汐和（近）惯性运动的周期分量（Up）和用于亚网格尺度运动(U)的拉格朗日随机分量（模型；LSM）的模型。LSM最初将基于高分辨率漂移数据。然后将开发基于大尺度速度信息（如积分时间尺度、涡动变化）和海岸地貌的LSM参数化。这使得轨迹模型可以应用于所有有足够大尺度流动信息的沿海地区。在圣巴巴拉（加利福尼亚州）和迈阿密（佛罗里达州）沿海地区，通过与原位漂船轨迹的比较，将对大范围动态参数（例如速度、涡度、涡流能量）的模型技能进行量化。RKFs给出了在时间t0在某个位置x0释放的粒子在t1到达位置x1的概率，将用模拟轨迹计算并作为预测沿海海洋运输或连通性的工具进行评估。随后确定RKFs在空间和时间上的变化方式和原因，将有助于提高对沿海环流的了解。

ABSTRACTOCE-0352187The PI's seek to determine optimal stochastic particle transport (i.e. Lagrangian) models for use in the coastal ocean which will provide a necessary link between the copious Eulerian coastal circulation data from high frequency (HF) radar systems, and transport information required by coastal resource managers tasked with identifying the fate of pollutants, larvae, and objects lost-at-sea. Specific research objectives are to (i) observe surface flow fields in two coastal regions using HF radar and high-resolution drifters, (ii) to develop accurate Lagrangian transport models to predict trajectories from the HF radar fields, (iii) to model trajectories and quantify their skill through comparisons with in-situ drifter tracks, and (iv) to compute redistribution kernel functions (RKFs), or connectivity matrices, and demonstrate their utility as simple probabilistic near-shore transport models. Existing trajectory models will be modified for consideration of flow patterns characteristic to the coastal ocean. A model with a large-scale mean component (U), a periodic component representing tidal and (near-) inertial motions (Up), and a Lagrangian stochastic component (model; LSM) for subgrid-scale motions (u) will be developed. The LSM will initially be based on high resolution drifter data. An LSM parameterization based on large-scale velocity information (e.g. integral time scales, eddy variance.) and coastal geomorphology will then be developed. This enables the trajectory models to be applied in all coastal regions with sufficient large-scale flow information. Model skill over a wide range of dynamic parameters (e.g. velocity, vorticity, eddy energy) will be quantified in the Santa Barbara (CA) and Miami (FL) coastal regions through comparisons with in-situ drifter tracks. RKFs, giving the probability that a particle released at some location x0 at time t0 will reach location x1 at t1, will be computed with modeled trajectories and evaluated as tools for predicting transport, or connectivity in the coastal ocean. An improved understanding of coastal circulation will result from subsequent identification of how and why RKFs vary spatially and temporally.