| The purpose of the research has been to provide a better understanding of the climatology, variability, mechanisms, physical processes, and predictability of the Intertropical Convergence Zone (ITCZ), one of the most prominent and important features of the tropical atmosphere. The statistical analyses show that there are a number of distinctly ITCZ regimes in terms of their structure and dynamics. Over the Indian and western Pacific oceans, the relationship between sea surface temperature and convection is strongly influenced by the large-scale atmospheric circulation.; The migration of the ITCZ depends strongly on a wide range of temporal scales. On the interannual time scale (17–84 months), the tropical ocean-atmosphere system is dominated with the El Niño/Southern Oscillation (ENSO) phenomena. The relationship and long-term predictability ENSO events and convective anomalies over Indonesia are examined using the Empirical Orthogonal Function (EOF) and Canonical Correlation Analysis (CCA) techniques. The results show that the models are potentially useful in predicting convective anomalies over Indonesia during boreal fall and winter, but have lower skills in spring and summer months caused by weakest Southern Oscillation and the so-called spring barrier.; The behavior of the northeast monsoon over the Indian Ocean and Indonesia during a normal and during an El Niño/Southern Oscillation (ENSO) year was compared using the Naval Research Laboratory/North Carolina State University (NRL/NCSU) model and the Fifth Generation of the Pennsylvania State University-National Center for Atmospheric Research Mesoscale Model (PSU-NCAR MM5). During the active monsoon in 1997, both models generally simulate the mean circulations up to 24 hours, but show rapid decline in the forecasts beyond that when the winds become stronger. In contrast, when the synoptic-scale dynamic forcing is strong during the 1998 ENSO event, the simulated mean flows are in good agreement with the analyses up to 48 hours. It is also found that the rates and distributions of oceanic and land mass rainfall are more realistically simulated by the NRL/NCSU model than by the PSU-NCAR MM5 for these two cases. The model errors can be attributed among others to the inaccuracies in the PBL parameterization and uncertainties in the initial conditions. |
