Development and Testing of a Unified Parameterization of Diabatic Heating - Proof of Concept

智力优势：首席研究员（PI）将开发一种新的数值模型物理参数化方法。典型的参数化是一个一维模型，它使用来自较大尺度模型的输入数据来估计被估计的物理实体（例如湍流通量）。PI将通过开发统一的参数化来绕过一维模型方法，该参数化将派生自查找表（LUT）并基于来自数值模型的输入变量。在这项研究中，首席研究员将专注于亚电网尺度的绝热加热。如果这种方法能够足够精确，那么模型的计算速度将会有显著的提高。PI将在2006年的选定月份使用标准的一维参数化方法在北美运行区域大气模拟系统（RAMS）。集成的非绝热加热速率将与输入变量和参数一起存储和使用，以创建一个统一的LUT，该LUT包括非绝热加热的每个原始标准参数化的效果总和。RAMS使用导出的LUT来模拟绝热加热，将在相同的时间段内进行整合，以确定所产生的气象场是否能以与原始模式运行相同的精度再现，但计算成本只是其中的一小部分。如果该技术被证明是准确的，PI将检查人为地在LUT中施加扰动时产生的模拟气象场的变化，以便评估结果对这些变化的敏感性。在研究结果中，将包括精确捕获输出空间维度所需的LUT的大小，这取决于使用统一LUT运行的区域模型在使用不同绝热加热项单独参数化的传统方法时重建原始结果的能力。更广泛的影响：该研究对科学界和社会具有广泛的影响。用lut取代天气和气候模型中现有的参数化，其成本只占原始参数化的一小部分，这将在模型运行速度方面带来革命性的改进。这项研究将直接有助于科罗拉多大学博尔德分校的一名研究生的教育。

Intellectual Merits: The Principal Investigator (PI) will develop a new methodology for physical parameterizations in numerical models. The typical parameterization is a one dimensional model that estimates the physical entity being estimated (e.g. turbulent fluxes) using as input data from the larger scale model. The PI will bypass the one dimensional model approach by developing a unified parameterization that would be derived from a look up table (LUT) and based on input variables from the numerical model. For this research, the Principal Investigator will concentrate on subgrid scale diabatic heating. If this approach can be made sufficiently accurate, there would be a significant gain in computational speed for the model. The PI will run the Regional Atmospheric Modeling System (RAMS) for selected months in 2006 over North America using the standard one-dimensional parameterizations for diabatic heating. The integrated diabatic heating rates will be stored and used along with the input variables and parameters to create a unified LUT that includes the sum of the effects from each original standard parameterization for the diabatic heating. RAMS, using the derived LUT to simulate the diabatic heating, will be integrated for the same time periods to determine if the resultant meteorological fields are reproduced with the same accuracy as in the original model run, but at a fraction of the computational cost. If the technique proves to be accurate, the PI will examine the variations in the simulated meteorological fields that result when perturbations are artificially imposed in the LUT in order to assess the sensitivity of the results to these variations. Among the results of the research will be the size of the LUT that is needed to accurately capture the dimensionality of the output space, as determined by the ability of the regional model run with the unified LUT to recreate the original results when the traditional approach of using separate parameterizations for the different diabatic heating terms is used. Broader Impacts: The research has broad impacts to the scientific community and to society. The replacement of existing parameterizations in weather and climate models with LUTs at a fraction of the cost of the original parameterization will provide a revolutionary improvement in model run speed. This research will directly contribute towards the education of one graduate student at the University of Colorado, Boulder.