A physical subgrid scale information exchange (PSIE) systemfor parameterization schemes in numerical weather prediction models

数值天气预报模式需要对基本微分方程组进行时空离散。通过引入数值网格，所有的次网格尺度过程都被截断，并且必须被参数化以包括它们对网格尺度平均值的影响，这些物理参数化显著地决定了数值天气预报模式的总体行为。现代物理参数化方案提供了所讨论的过程的详细信息，但它们仅将这些信息的一部分作为平均值提供给网格尺度。不同的参数化方案之间通常不直接传递次网格尺度信息。这种通信的唯一方式是通过平均模型变量。由于网格尺度平均和对所选预测变量集的限制，这导致不可避免的信息丢失。为了表征和克服这种信息损失，在拟议的研究项目中，我们计划开发一个物理次网格尺度信息交换（PSIE）系统的数值天气预报模式的参数化方案。与次网格尺度过程相关的信息将在参数化之间直接交换。预期这种处理产生天气预报模型的整体预报质量的明显改善。由于预报变量和物理参数化的选择取决于托管天气预报模式，我们将使用两种不同的托管模式进行比较：德国气象局的COSMO模式和美国国家大气研究中心和国家海洋和大气管理局开发的WRF模式。

Numerical weather prediction models have to solve the basic differential equations in aspatially and temporally discretized form. By introducing a numerical grid, all processes of subgridscales are cut off and have to be parameterized to include their influence on the grid scale averages.These physical parameterizations significantly determine the total behavior of the numerical weather prediction model. Contemporary physical parameterization schemes purvey detailed information of the processes in question, but they only provide a part of this information as an average to the grid scale. Subgrid scale information is usually not directly communicated between the different parameterization schemes. The only way for this communication is via the averaged model variables. This causes an inevitable loss of information due to the grid scale averaging and the restriction to the chosen set of prognostic variables. To characterize and to overcome this loss of information, in the proposed research project we plan to develop a physical subgrid scale information exchange (PSIE) system for the parameterization schemes of a numerical weather prediction model. Information related to subgrid scale processes will be exchanged directly between the parameterizations. It is expected that this treatment yields a distinct improvement of the overall forecast quality of the weather prediction model. Since the choice of the prognostic variables and the physical parameterizations depend on the hosting weather prediction model model, we will use two different hosting models for comparison: the COSMO model of the German Meteorological Service and the WRF model developed by the National Center for Atmospheric Research and the National Oceanic and Atmospheric Administration, USA.