中文摘要：

由格子点的二个版本的模仿的出去的 longwave 放射(OLR ) 产量大气的一般发行量模型(GAMIL ) 被分析在 Madden-Julian 摆动(MJO ) 和另外的热带波浪的模拟上在云微视物理学和对流 parameterization 计划估计改进的影响。wavenumber 频率光谱分析被使用孤立对流地联合的赤道的波浪的主导的模式包括 MJO，凯尔文，赤道的 Rossby (嗯) ，混合 Rossby 严肃(MRG ) ，和 inertio 严肃(IG ) 飘动。GAMIL 的不同版本的表演当模特儿(版本 1.0 (GAMIL1.0 ) 并且版本 2.0 (GAMIL2.0 )) 被在 GAMIL1.0， GAMIL2.0，和观察数据之中比较这些波浪的力量光谱分布评估。GAMIL1.0 显示出一个弱 MJO 信号，与在 wavenumbers 独立发生的最大的可变性 1 和 4 而非是专注于 wavenumbers 13，建议那 GAMIL1.0 不能有效地捕获 intraseasonal 可变性。然而， GAMIL2.0 能有效地复制对称、反对称的波浪，和波浪与观察数据一致的 MJO ，凯尔文，和 MRG 的重要系列，显示 GAMIL2.0 的能力被改进云微视物理学和对流 parameterization 计划并且暗示如此的改进是关键的推进改进这个模型的表演提高模仿 MJO 和另外的热带波浪。

英文摘要：

Simulated outgoing longwave radiation （OLR） outputs by two versions of the grid-point atmospheric general circulation model （GAMIL） were analyzed to assess the influences of improvements in cloud microphysics and convective parameterization schemes on the simulation of the Madden-Julian oscillation （MJO） and other tropical waves. The wavenumber-frequency spectral analysis was applied to isolate dominant modes of convectively coupled equatorial waves, including the M30, Kelvin, equatorial Rossby （ER）, mixed Rossby-gravity （MRG）, and inertio-gravity （1G） waves. The performances of different versions of the GAMIL model （version 1.0 （GAMIL1.0） and version 2.0 （GAMIL2.0）） were evalu- ated by comparing the power spectrum distributions of these waves among GAMIL 1.0, GAMIL2.0, and observational data. GAMIL1.0 shows a weak MJO signal, with the maximum variability occurring separately at wavenumbers 1 and 4 rather than being concentrated on wavenumbers 1-3, suggesting that GAMILI.0 could not effectively capture the intraseasonal variability. However, GAMIL2.0 is able to effectively reproduce both the symmetric and anti-symmetric waves, and the significant spectra of the MJO, Kelvin, and MRG waves are in agreement with observational data, indicating that the ability of GAMIL2.0 to simulate the MJO and other tropical waves is enhanced by improving the cloud microphysics and convective parameterization schemes and implying that such improvements are crucial to further improving this model＇s performance.