Large-Scale Parallel Computing of Cloud Resolving Storm Simulator
Large-Scale Parallel Computing of Cloud Resolving Storm Simulator
复制标题
云解析风暴模拟器大规模并行计算
DOI：
--
复制
发表时间：
2002
期刊：
影响因子：
--
通讯作者：
Atsushi Sakakibara
中科院分区：
文献类型：
--
作者：
K. Tsuboki;Atsushi Sakakibara
强雷暴是由强对流云组成的。为了模拟这种类型的风暴，一个非常精细的网格系统是必要的，以解决单个对流云在一个大的域。由于对流云是一个高度复杂的云动力学和微物理系统，因此需要详细地描述云的物理过程和流体动力学。计算需要大量的内存和大规模的并行计算。对于这种类型的计算，我们已经开发了一个云解析数值模型，这被命名为云解析风暴模拟器（CReSS）。在本文中，我们将详细描述CReSS的基本配方和特点。我们还展示了一些使用CReSS的大规模并行计算获得的风暴的数值实验结果。
A sever thunderstorm is composed of strong convective clouds. In order to perform a simulation of this type of storms, a very fine-grid system is necessary to resolve individual convective clouds within a large domain. Since convective clouds are highly complicated systems of the cloud dynamics and microphysics, it is required to formulate detailed cloud physical processes as well as the fluid dynamics. A huge memory and large-scale parallel computing are necessary for the computation. For this type of computations, we have developed a cloud resolving numerical model which was named the Cloud Resolving Storm Simulator (CReSS). In this paper, we will describe the basic formulations and characteristics of CReSS in detail. We also show some results of numerical experiments of storms obtained by a large-scale parallel computation using CReSS.