Resolvable-and Subgrid-Scale Measurements in the Atmospheric Surface Layer

摘要/Wyngaard ATM-9704740大涡模拟（LES）已成为微气象研究的首选模拟方法。然而，由于数值分辨率不足和严重依赖亚网格尺度（SGS）模型的精度，其在表层的性能受到影响。直到最近，很少有人注意到表层的这些缺陷，可能有两个原因。首先，实验上的限制使得很难产生对LES进行彻底测试的数据。例如，没有对LES预测的可分辨尺度场的直接测量报告。其次，传统的SGS模型在表层的弱点还没有得到详细的记录。本文建议在大气表层测量可分辨尺度和亚网格尺度的大气湍流，以测试和改进LES的SGS关闭。最近的研究表明，在地表附近广泛使用的Smagorinsky SGS模型存在主要缺陷。提出了一种新的实验技术，即利用传感器的横向阵列来测量大气表层的二维过滤湍流变量。这里提出的初步研究证明了这项技术的潜力，将在拟议的研究中加以扩展。在近中性到高对流的ABL状态范围内，将在表层测量可解析变量和SGS变量，并对其进行分析，以详细研究可解析变量和亚网格尺度动力学。所得数据和新的物理见解将与使用直接数值模拟和LES数据的并行程序相结合，为改进的表层湍流SGS模型的发展提供必要的动力。该计划是宾夕法尼亚州立大学一个多学科研究小组协调努力的一部分，旨在通过结合数值模拟和现场测量来分析和改进SGS模型。本研究将为遥感技术的发展提供基准数据，用于测量表层过滤湍流。

Abstract/Wyngaard ATM-9704740 Large-eddy simulation (LES) has become t he modeling method of choice in micrometeorological research. Nonetheless its performance in the surface layer suffers because of inadequate numerical resolution and heavy reliance on the precision of the subgrid-scale (SGS) model there. Until recently, little attention has been focused on these deficiencies in the surface layer, probably for two reasons. First, experimental limitations make it difficult to produce data for thorough testing of LES. No direct measurements of the resolvable-scale fields that LES predicts have been reported, for example. Second, the weaknesses of the traditional SGS models in the surface layer have not been documented in much detail. Measurements of resolvable- and subgrid-scale atmospheric turbulence in the atmospheric surface layer for the purpose of testing and improving SGS closures for LES are proposed here. Recent studies indicate major weaknesses of the widely used Smagorinsky SGS model near the surface. A new experimental technique is proposed whereby a lateral array of sensors is used to measure two-dimensionally filtered turbulence variables in the atmospheric surface layer. The preliminary studies presented here, which demonstrate the potential of the technique, will be extended in the proposed studies. Resolvable and SGS variables will be measured in the surface layer over a range of ABL states, from nearly neutral to highly convective, and shall be analyzed to study resolvable and subgrid-scale dynamics in detail. The resulting data and new physical insight will be coupled with parallel program using direct numerical simulation and LES data to provide the necessary impetus for the development of improved SGS models for surface-layer turbulence. The proposed program is part of a coordinated effort in a multi- disciplinary research group at the Penn State University directed at analysis and improvement of SGS modeling through combined numerical simulation and field measurement. The present studies will provide benchmark data for the development of remote sensing techniques for the measurement of filtered turbulence in the surface layer.