资源环境科技发展态势分析平台

A regional model is used to quantify the influence of the extratropics on simulated tropical intraseasonal variability. The Weather Research and Forecasting (WRF) model is run in tropical channel mode with the boundaries at 30N and S constrained to 6-hourly reanalysis data. Experiments with modified boundary conditions are carried out in which intraseasonal (20-100 days) timescales are removed, or in which only the annual and diurnal cycles are retained. Twin runs are used to give an objective measure of the boundary-independant component of the variance in each case. The model captures MJO-like propagating structures and shows greater zonal-wind variance in runs with full boundary conditions. Comparison between experiments indicates that about half the intraseasonal variance can be attributed to boundary influence, and specifically to the presence of an intraseasonal extratropical signal. This signal is associated with stronger correlations between low-level zonal wind precursors in the Pacific sector and Indian Ocean convective events. Temporal coherence between MJO events in the model and the observations is analysed by defining four phases based on convectively coupled signals in the low-level zonal wind. The model can only match observed events above the level of chance when intraseasonal boundary information is provided. Results are analysed in terms of 'primary' and 'successive' events. Although the model hindcast skill is generally poor, it is better for successive events.