Microscale and Mesoscale Structures in Convective Marine Boundary Layers

9419927这笔赠款将通过对特殊领域数据的观测研究和理论模型模拟，研究对流海洋边界层的动力和热力学性质和特征。这些云层顶端的边界层是对冷大陆气团的响应，这些冷气团在相对温暖的水中平流，导致对流混合，形成位于大气层两公里内的大型多边形中尺度细胞对流阵列。这种对流模式可能以二维条带或街道的形式存在，但通常(下游)会分解成三维单元。在过去的三十年里，通过各种手段的测量积累了丰富的大气对流研究观测数据，包括雷达、飞机和特殊野外观测计划中的卫星。从这些数据中获得了描述行星边界层(PBL)结构和动力学的统计参数。这些参数通常以时间或空间平均值的形式从有限长度的几个(通常从一个)记录中计算，而理论研究采用集合平均值，该集合平均值假定随机函数在无限数量的实现上的平均值。只有当潜在的随机函数是平稳的(齐次的)时，集合平均才能用足够大的时间或空间分段的平均来近似。目前的气象研究往往由于数据的非平稳性或不均质性以及平均距离不足而导致统计上不可靠的推断。由于这些原因，在该项目下，将采用最近开发的综合统计方法重新审查在几个外地方案期间收集的数据集，该方法提供对统计参数及其统计可靠性的估计。大涡模拟(LES)是从理论上理解大气对流复杂流体力学的又一有力工具。再次存在获得统计上可靠的结果的问题，本项目将通过将上述统计方法从随机过程扩展到随机场并使用蒙特卡罗技术近似总体平均来解决这一问题。还将对边界层中的中尺度对流进行新的大涡模拟，并对大涡模式资料进行适当的统计分析。重点将放在含有冰粒和降雪的对流PBL的物理性质上。模型结果将与来自现场项目的统计可靠的结果进行比较。

9419927 Agee This grant will address the dynamic and thermodynamic properties and characteristics of convective marine boundary layers, through both observational study of special field data and theoretical model simulations. These cloud-topped boundary layers form in response to cold continental air masses that advect out over relatively warmer water, resulting in convective mixing and the formation of large polygonal arrays of mesoscale cellular convection that reside within the first one of two kilometers of the atmosphere. This convective pattern may exist as two- dimensional bands or streets, but often (downstream) resolves itself into three dimensional cells. During the past three decades a wealth of observational data for atmospheric convection studies has been accumulated from measurements by various means, including radar, aircraft, and satellites in special field observation programs. From the data, statistical parameters were obtained describing the structure and dynamics of the planetary boundary layers (PBL). These parameters are commonly computed in the form of temporal or spatial averages from a few (often from one) records of limited length, while theoretical studies operate with ensemble averages which assume averaging over an infinite number of realizations of random functions. Ensemble averages can be approximated by the averages taken over sufficiently large time or space segments only if the underlying random function is stationary (homogeneous). Current meteorological studies often result in statistically unreliable inference because of nonstationarity or inhomogeneity of data and inadequate averaging distances. For these reasons, under this project, data sets collected during several field programs will be revisited by employing a recently developed comprehensive statistical method that provides estimates of both statistical parameters and their statistical reliability. Large eddy simulation (LES) modeling is anoth er powerful tool that helps provide a theoretical understanding of the intricate fluid mechanics of atmospheric convection. Once again there is a problem of obtaining statistically reliable results, which will be addressed in this project by extending the above statistical method from random processes to random fields, and by approximating ensemble averages using Monte Carlo techniques. New LES model development of mesoscale convection in PBLs and proper statistical analysis of the LES model data will also be performed. The focus will be on the physical properties of convective PBLs with ice particles and snow precipitation. Model results will be compared with statistically reliable findings from field programs.