Evaluation and Improvement of Microphysical Parameterizations in Mesoscale Models

高分辨率天气预报模式的使用已成为区域天气和降水业务预报的基石。 但是，即使在模式分辨率和物理学方面取得了稳步进展，降水预报的改进也很缓慢。 云和降水过程的模式体微物理参数化（BMP），即湿过程的模式描述，存在着明显的不足。 这些缺陷的根本原因是BMP所依据的许多假设存在相当大的不确定性。 提高BMP性能的一种途径是将微物理过程和来自模型模拟的预测云/降水分布与现场（主要是空中）和遥感（例如，雷达观测。 此外，至关重要的是，微物理测量与风、温度和湿度的观测同时进行，这样模拟的微物理误差就可以与其他预测场的误差隔离开来。 为了解决这些问题，华盛顿大学的研究人员发起了一项名为“通过观测验证实验（IMPROVE）改进微物理参数化”的研究，以获得将当前预报和研究模型中的云和降水过程与各种天气系统的详细测量和观测进行比较所需的观测结果。 进行了两项实地研究：华盛顿近海锋面研究（IMPROVE-1），研究2001年1月4日至2月14日接近华盛顿海岸的锋面系统;俄勒冈州小瀑布地形研究（IMPROVE-2），研究2001年11月26日至12月22日俄勒冈州小瀑布上空云和降水的地形调制。 利用一系列综合观测平台，这两项实地研究都非常成功地获得了用于评估中尺度模式中BMP性能的数据，其中26个密集观测期（IOP）涵盖了各种各样的锋面和地形降水系统。 对这些数据的分析表明，MM 5模式中最复杂的微物理方案存在重大问题，包括迎风坡和山顶的积雪量过多，背风坡的积雪过多，迎风坡下部的云液态水太多，山顶上的云液态水太少，有问题的积雪尺寸分布，以及中层不切实际的<$球量。 这项研究将确定这些和其他微物理问题是否在其他情况下被揭示出来。 具体而言，将分析来自额外的IMPROVE IOP的观测数据，以确定在观测情况下导致云/降水发展的物理过程，然后与中尺度模型模拟进行比较，分辨率约为1公里。 这些比较将为修改BMP提供基础，以便更好地代表云和降水的发展。 此外，IMPROVE期间NOAA P3飞机飞行的双多普勒雷达数据将用于评估模型中山波和其他关键中尺度结构的保真度。 最后，根据上述评价，将对IMPROVE期间研究的各种风暴系统以及华盛顿大学实时MM 5/WRF区域预报系统的日常业务预报运行中的模型湿物理学修改进行评价。 关于智力价值，主要研究人员将分析可能是最全面的数据集在现有的处理流动和潮湿的物理地形障碍。 这项研究将提供最好的评估中尺度模式结构和潮湿的物理地形的保真度，并应导致天气预报模式中的潮湿过程参数化的改进。 这项研究有可能为社会带来广泛的利益。 查明和纠正天气预报模型中潮湿物理学的不足之处，应能增进对云和降水的了解和预测，并产生相应的社会和经济效益。 该项目所产生的物理理解和模型改进将广泛传播，供业务预报中心和其他团体使用。 这一努力也将有实质性的教育效益，创造一组研究生在这一关键领域的知识，并将暴露一些本科生这一重要的研究课题。

The use of high-resolution weather prediction models has become the cornerstone of operational forecasting of regional weather and precipitation. But even with steady advances in model resolution and physics, precipitation forecasts have been slow to improve. It has become evident that there are substantial deficiencies in model bulk microphysical parameterizations (BMP) of cloud and precipitation processes, that is, the model descriptions of moist processes. Underlying these deficiencies are considerable uncertainties in many of the assumptions on which BMPs are based. One avenue to improve the performance of a BMP is to compare microphysical processes and predicted cloud/precipitation distributions from model simulations with in situ (mainly airborne) and remotely sensed (e.g., radar) observations. In addition, it is critically important that the microphysical measurements be obtained concurrently with observations of wind, temperature and humidity, so that errors in the simulated microphysics can be isolated from errors in other predicted fields. In response to these problems, UW researchers initiated a study entitled Improvement of Microphysical PaRameterization Through Observational Verification Experiment (IMPROVE) to acquire the observations required to compare cloud and precipitation processes in current forecast and research models with detailed measurements and observations from a variety of weather systems. Two field studies were conducted: the Washington Offshore Frontal Study (IMPROVE-1), which examined frontal systems as they approached the Washington coast from 4 January to 14 February 2001; and, the Oregon Cascades Orographic Study (IMPROVE-2), which examined the orographic modulation of clouds and precipitation across the Oregon Cascades between 26 November and 22 December 2001. Making use of a comprehensive array of observing platforms, both field studies were highly successful in obtaining data for evaluating the performance of BMPs in mesoscale models, with twenty-six Intensive Observing Periods (IOPs) encompassing a wide variety of frontal and orographic precipitation systems. Analysis of this data has documented significant problems with the most sophisticated microphysics scheme in the MM5 model, including excessive snow amounts over the windward slopes and mountain crest, excessive snow blow-over to the lee slopes, too much cloud liquid water over the lower windward slopes and too little over the crest, problematic snow size distributions, and unrealistic graupel amounts at mid-levels. This research will determine if these and other microphysical problems are revealed in other cases. Specifically, observational data from additional IMPROVE IOPs will be analyzed to ascertain the physical processes leading to the development of clouds/precipitation in the observed cases, followed by a comparison with mesoscale model simulations down to resolutions of approximately 1 km. These comparisons will provide the basis for modifications to BMPs in order to better represent the development of clouds and precipitation. In addition, the dual-Doppler radar data from NOAA P3 aircraft flights during IMPROVE will be used to evaluate the fidelity of mountain waves and other key mesoscale structures in the model. Finally, based on the above evaluations, modifications in the model moist physics will be evaluated for a wide variety of storm systems studied during IMPROVE, as well as in daily operational forecast runs of the University of Washington's real-time MM5/WRF regional forecast system. Regarding intellectual merit, the Principal Investigators will analyze probably the most comprehensive data set in existence dealing with the flow and moist physics over a topographic barrier. This study will provide the best evaluation to date of the fidelity of mesoscale model structures and moist physics over terrain and should lead to improved moist process parameterizations in weather prediction models. The research has the potential for broad benefits to society. The identification and correction of deficiencies in the moist physics of weather forecast models should result in improved understanding and prediction of clouds and precipitation, with the attendant societal and economic benefits. The physical understanding and model improvements resulting from this project will be widely disseminated for use by operational forecasting centers and other groups. This effort will also have substantial educational benefits, creating a group of graduate students knowledgeable in this critical area, and will expose a number of undergraduates to this important research topic.