The role of the Indian Ocean in climate variability was overlooked until recently. This was basically because an independent basin-scale ocean-atmosphere coupled mode was not recognized. Following recent discovery of the Indian Ocean dipole (IOD) mode and its impact on global climate, the need of a long-term monitoring of the Indian Ocean (Figure 1) is evolving as a major issue of CLIVAR/WCRP. In addition to the IOD, this monitoring system will also help to understand the intraseasonal disturbances, annual/monsoon, interannual, decadal scales variations and long-term warming trend, cross-equatorial overturning cells, Indonesian Throughflow, western and eastern boundary currents and ocean domes etc. The Honolulu meeting was arranged to discuss the design of the observational network based on numerical results from linear to complex coupled general circulation models.

A few preliminary studies discussed excellent results from the available mooring arrays initiated by JAMSTEC and a few agencies from India and US. In the equatorial Indian Ocean interesting biweekly signals is observed in meridional surface current related to the oceanic Yanai-Maruiyma wave (mixed gravity-Rossby wave). Several Ocean General Circulation Models (OGCMs) successfully simulated the phase and energy propagation of this wave and provided data for better understanding of its mechanism. Active intraseasonal and semiannual signals in the equatorial zonal upper-ocean current were also revealed by the mooring arrays.

The tropical Indian Ocean is known to be a breeding ground for intraseasonal disturbances. Several presentations showed their influence on the IOD, El Nino Southern Oscillation (ENSO), and monsoon. Predictability of these intraseasonal signals is very challenging. Coupled model studies showed the importance of simulating a realistic mean state for predicting the eastward propagation and strength of these intraseasonal disturbances. Potential impact of mixed layer diurnal cycle on tropical convections and intraseasonal disturbances was also discussed.

The interactive relationship between ENSO and IOD has been largely debated in the community. During this workshop, many CGCMs results clearly showed the existence of IOD events. In particular, carefully designed experiments using the SINTEX-F demonstrated the air-sea coupling of IOD in the absence of ENSO. Using 200-yr data from the control experiments, it was shown that the heat content anomalies in the tropical Indian Ocean are responsible for the decadal variations of IOD occurrences. Effects of IOD and ENSO on the Indonesian Throughflow and the biennial character of the Indian monsoon also were presented. Several studies showed significant impact of the Indian Ocean Sea Surface Temperature (SST) changes on the rainfall variability of the southern Africa.

Abundance of rainfall gives rise to barrier layers in several parts of the basin. A process study using the SINTEX-F coupled model showed the significant effect on the barrier layer in the spring SST warming of southeastern Arabian Sea and early monsoon onset. Several presentations showed the importance of correctly simulating the inter-basin water exchanges particularly between the Bay of Bengal and the Arabian Sea. It is suggested that salinity and solar radiation penetration into the shallow mixed layer need to be measured. Saline water from Red Sea and water input from the Pacific via the Indonesia Throughflow also are important for the Indian Ocean variability.

Compared to ENSO, current CGCMs have greater difficulties in simulating and predicting Indian Ocean climate signals. One presentation using the FRCGC SINTEX-F CGCM showed skillful scores of up to 4 months lead for IOD prediction. The long-term observation system will certainly help us to understand and predict Indian Ocean climate changes. We are at the dawn of an active period of Indian Ocean climate research.