Refereed papers on most data sets collected during TOGA-COARE, a major field experiment conducted from November 11992 to February 28 1993 in the equatorial western Pacific, are now published or in press. A review of the present status of COARE research has been prepared recently (Godfrey et al., 1996). It is an interim review, since the full "payload" of COARE (in terms of new, observationally-based algorithms to improve the performance of General Circulation Models) is not likely to be delivered till about the year 2000. There are a number of exciting new results from COARE - e.g. the elucidation of the role of 4-day waves in advecting dry air into the COARE region (Numaguti, 1995), or the new understanding of westward propagation of deep convective events (Chen and Houze, 1996). However, this short talk, drawn from Godfrey et al. (1996), will concentrate on those aspects of COARE which in the author's opinion are most likely to result in major modifications of coupled climate models. These relate to a better understanding of the physics of IntraSeasonal Oscillation (ISO) episodes - fluctuations in rain, wind and SST on a 30-60 day timescale that play a major role in ENSO events.
Organization of COARE:
During the Intensive Observation Period (IOP) of November 1992 through February 1993, observations were collected on all the platforms shown in Fig. 1.The top panel shows the outer sounding array, or OSA, and the Intensive Flux Array or IFA, both with four soundings per day. The OSA and IFA were designed for use in atmospheric budget studies. Experience in earlier experiments showed that data from such arrays also provided valuable consistency checks on aircraft observations. Seven Integrated Sounding System (ISS) sites were located at Nauru, Kapingamarangi, Manus and Kavieng, and on three ships - Kexue 1, Shiyan 3, and Moana Wave - to provide more detailed recording of convective events than was possible from conventional soundings. Two 5-cm wavelength Doppler weather radars were deployed on ships in COARE. One was the TOGA radar aboard the Xiangyanghong #5; the other was the MIT radar aboard the Vickers. At least one of the ships was present in the WA throughout all but a few days of COARE. These radars show the pattern of rain over the ocean in great detail and provide estimates of the rain rates. The two shipborne Doppler radars were supplemented for a short time in November by an operational radar aboard the Keifu Maru. A pair of X-band (3-cm wavelength) Doppler radars operated at Manus Island for two and a half months. Several moorings are also indicated, including a number of ADCP and Proteus moorings for measuring currents, and the IMET mooring with comprehensive and redundant meteorological instrumentation. These moorings were in addition to western Pacific enhancements of the regular Pacific-wide TAO array. Several research vessels devoted their efforts to documenting changes in the upper ocean, mostly near the IFA.
Largescale circulation during COARE
Figure 2a shows a Hovmoller diagram of monthly mean zonal wind anomalies and SST anomalies along the equatorial Pacific, from the TOGA-TAO array (Hayes et al., 1991), for the period 1987-1993. The well-known relationship of SSTA's in the east
