Collaborative Research: Observational and Numerical Modeling Studies of Rain Microphysics

自1988年以来，美国国家气象局已经运行了一个由大约160个先进的多普勒天气雷达组成的网络，用于监测恶劣天气并发布预警，例如龙卷风、山洪暴发等。这些雷达的上一次重大升级发生在2013年，涉及双极化能力，大大增强了探测山洪风暴和登陆飓风极端降雨的能力。该项目旨在利用先进的仪器测量单个雨滴的特性，如大小、形状、浓度和下降速度，从而提高这种雷达对降雨的测量精度。根据对不同降雨强度和类型的这些特性的了解，雷达估计降雨率的算法可以比迄今为止更准确地发展。与此同时，利用降雨形成和演变的详细微物理的数值模型正在与雷达观测进行比较，以验证数值模型中的物理表征和假设确实是正确的。这是很困难的，因为地面的大雨起源于风暴中较冷温度下的许多高空来源，并且随着雨滴到达地面，它以一种复杂的方式演变。因此，将雷达测量数据与地面降雨特性和数值模型相结合是本研究的重要组成部分，这可能会提高国家气象局发布的洪水风暴警报的准确性。该项目的主要科学目标是通过双极化雷达对DSD矩的检索和一维（1D）模型的协同组合，获得对水滴尺寸分布（DSD）演变的微物理过程的更深入的理解——通过扫描极化c波段ARMOR雷达和由阿拉巴马大学亨茨维尔分校（UAH）运营的移动综合剖面系统（MIPS），对观测到的情况进行矩和过程速率的预测。模拟基于两个新的数值模型，(i)俄克拉荷马大学开发的云粒子模型（CPM），它明确地解释了暖雨微物理过程中所有云粒子的演变，（ii）德国气象局开发的一种新的蒙特卡罗微物理模型（McSnow），它基于“超级粒子”的物理特性模拟了冰、混合相和雨的演变。三种不同的仪器将用于测量和表征0.1-8毫米范围内的dsd，精度很高（气象粒子光谱仪，2d视频disdrometer和降水发生传感器系统）。同时，仪器上方的体积将被双极装甲雷达和MIPS扫描。这两种粒子模型将在一维中运行，DSD的演变将与地面测量和雷达剖面进行比较，以推断形成DSD的主要微物理过程。融化水平以上的冰过程与融化水平以下的雨过程之间的耦合需要更好地理解。mcsnow预测的双pol变量从明亮波段以上到地表高度的剖面和斜率将与雷达观测结果进行比较，以推断主要过程和降雨类型。地表测量将成为模型预测的约束条件。在几乎所有的碰撞过程模型中，一个中心假设是雨滴的最终速度只取决于雨滴的质量。然而，最近对落体速度分布的一些观测表明，在湍流条件下，毫米大小的落体会明显偏离Gunn-Kinzer终端速度方程，这需要得到证实。显式云粒子模型将用于预测湍流引起的坠落速度偏差（均值和方差）是否对碰撞过程和随后的DSD演化有重大影响。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Since 1988, the US National Weather Service has operated a network of around 160 advanced Doppler weather radars for monitoring severe weather and issue warnings, for example, of tornadoes, flash-floods, etc. The last major upgrade of these radars occurred in 2013 involving dual-polarization capability which greatly enhances the ability to detect flash-flood producing storms and extreme rainfall from land-falling hurricanes. This project seeks to improve the measurement accuracy of rainfall by such radars using advanced instruments that measure the properties of individual rain drops such as size, shape, concentration and fall speeds. From knowledge of these properties in different rain intensities and types, the algorithms for radar estimation of rainfall rates can be developed with greater accuracy than possible hitherto. In parallel, numerical models that use detailed microphysics of rain formation and evolution are being compared against radar observations to verify that the physical representation and assumptions in the numerical models are indeed correct. This is difficult because heavy rain at the surface originates from many sources aloft at colder temperatures within the storm and it evolves in a complex manner as the drops reach the surface. Thus, the integration of radar measurements with surface rain properties and numerical models is an essential component of this research which is likely to result in improved accuracy of warnings of flood-producing storms issued by the National Weather Service. The main scientific goal of this project is to obtain a deeper understanding of microphysical processes governing the evolution of drop size distributions (DSDs) using a synergistic combination of dual-polarization radar retrievals of DSD moments and one-dimensional (1D) model-predictions of moments and process rates for well-observed cases by the scanning polarimetric C-band ARMOR radar and the Mobile Integrated Profiling System (MIPS) operated by the University of Alabama at Huntsville (UAH). The simulations are based on two new numerical models, (i) the cloud particle model (CPM) developed at the University of Oklahoma which explicitly accounts for the evolution of all the cloud particles under warm rain microphysical processes, and (ii) a novel Monte-Carlo microphysics model (McSnow) developed by the German Weather Service that simulates the evolution of ice, mixed phase and rain based on the physical properties of "super-particles". Three different instruments will be used to measure and characterize the DSDs over the entire range of sizes from 0.1-8 mm with good accuracy (Meteorological Particle Spectrometer, 2D-video disdrometer, and Precipitation Occurrence Sensor System). Simultaneously, the volume above the instruments will be scanned by the dual-pol ARMOR radar as well as MIPS. The two particle models will be run in 1D and the DSD evolution will be compared against surface measurements and radar profiles to infer the dominant microphysical processes that shape the DSD. The coupling between ice processes above the melting level to rain processes below the melting level to the surface needs to be better understood. McSnow-predicted profiles and slopes of dual-pol variables with height from above the bright-band to the surface will be compared with radar observations to infer the dominant processes as well as rain types. Surface measurements will serve as constraints to the model predictions.One central assumption in nearly all collisional process modeling is that raindrops fall at terminal velocity depending only on the drop mass. However, some recent observations of fall speed distributions show that under turbulent conditions, mm-sized drops can deviate significantly from the Gunn-Kinzer terminal velocity equation which needs to be confirmed. The explicit cloud particle model will be used to predict if there is a significant impact of turbulence-induced fall speed deviations (mean and variance) on collisional processes and subsequently on DSD evolution.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.