Collaborative Research: Data Assimilation of Dual Doppler Lidar Observations of the Urban Boundary Layer

晴空中的空气运动可以通过多普勒激光雷达探测到，它主要看到气溶胶颗粒在空气中移动。因此，扫描多普勒激光雷达系统可以表征一部分大气边界层（接近地球表面）的空气运动，但它只能看到运动的一个组成部分，即激光雷达光束方向上的径向部分。一种被称为数据同化的技术已经被开发出来，通过单个多普勒激光雷达的测量来约束流的数值模型来确定三维风场。当使用一段时间内的重复测量时，同化方案可以产生除了直接测量的信息之外的附加信息，包括速度的横梁分量。然而，推导出的风场在一定程度上依赖于模型中的假设，因此尚未证明该方法能够提供真实的风场重建。在这个项目中，研究人员将使用两个扫描多普勒激光雷达从不同的方向观察风场的相同部分，因此将直接测量风矢量的两个组成部分。这些测量将以两种方式使用：一次只使用一个激光雷达对数据同化所得的结果进行测试，然后由另一个激光雷达提供验证；将数据同化技术扩展到使用两个激光雷达，从而获得更好的风场重建。重建的风场将用于研究湍流相干结构及其在城市边界层中的作用，特别是因为它们可能影响污染物的输送和扩散。由于激光雷达同时提供风散射和后向散射信息，数据集可以约束模型对边界层中气溶胶传输和扩散的预测。这些检索技术将为研究湍流的结构和统计提供一个重要的工具。本研究的预期成果包括：验证（和可能的改进）应用于单激光雷达观测的数据同化技术，单激光雷达观测比双激光雷达观测更常见，获得成本更低；关于大气边界层湍流结构的性质及其对弥散的贡献的新信息；提高了预测分散的能力。研究结果将提高监测和预测城市环境中扩散的能力，可能对公共安全有重要的好处。

Air motions in the clear air can be detected by Doppler lidar, which sees primarily aerosol particles moving with the air. A scanning Doppler lidar system can therefore characterize the air motions in a portion of the atmospheric boundary layer (near the surface of the earth), but it only sees one component of the motion, the radial component in the direction of the lidar beam. A technique called data assimilation has been developed to determine the three-dimensional wind field from a single Doppler lidar by using the measurements to constrain a numerical model of the flow. When repeated measurements over time are used, assimilation schemes can produce additional information in addition to that measured directly, including cross-beam components of the velocity. However, the deduced wind fields are to some extent dependent on assumptions in the model, so it has not been demonstrated that this approach provides realistic reconstruction of the wind fields. In this project, the investigators will use two scanning Doppler lidars to observe the same parts of the wind fields from different directions and so will have direct measurements of two components of the wind vector. These measurements will be used in two ways: to test the results obtained by data assimilation using only one lidar at a time, for which the other lidar then provides verification, and to extend the data-assimilation technique to use two lidars and thus obtain a still better reconstruction of the wind field. The reconstructed wind fields will then be used to study turbulent coherent structures and their roles in the urban boundary layer, especially as they may influence the transport and dispersion of pollutants. Because the lidars provide both wind and backscatter information, the data set can constrain the model predictions of transport and dispersion of aerosols in the boundary layer. These retrieval techniques will thus provide an important tool for studying the structures and statistics of turbulence. The expected outcomes of this research include: verification (and possible improvement) of the data-assimilation technique as applied to single-lidar observations, which are far more common and less expensive to obtain than dual-lidar observations; new information on the nature of turbulent structures in the atmospheric boundary layer and their contributions to dispersion; and improved ability to predict dispersion. The results should lead to better capabilities for monitoring and predicting dispersion in an urban setting, with possible important benefits to public safety.