A Parallelization Scheme for Atmospheric General Circulation Model
大气环流数值模式的一种并行化方案

An efficient scheme suitable to convert a serial numerical model to parallelized one is summed up based on the general parallel computing theory and the common characteristics of atmosphere models, and the associated problems needing to be dealt with carefully are also discussed. The proposed scheme includes four steps:model structure analysis, time cost analysis, data decomposition with communication design and uniform debugging. These steps can help the numerical model programmer to complete the model's parallelization effectively. As an example, this scheme is used in the parallelization of the climate Spectral Atmosphere Model (SAMIL) with the resolution of R42L26 recently developed at the State Key Laboratory of Numerical Modeling Atmospheric Sciences and Geophysical Fluid Dynamics, Institute of Atmospheric Physics of Chinese Academy Sciences (LASG/IAP). According to the calculating structure in this model, different parallel skills are adopted in the Gaussian-grid and spectral spaces respectively. First, the radiation process which costs most according to the time-cost analysis is parallelized in OPENMP method since it is carried through the vertical column without data communication on other horizontal grids. Then in order to get better speedup, MPI is used to parallelize the physical grid-space calculating including all physical processes by dividing latitudes into several sections based on the model structure analysis. Moreover, some other processes such as time-integral and spectral-space calculations are also parallelized to improve the whole performance. Though this procedure is simple and can be implemented easily, the tested performances show that the parallel efficiency is acceptable.