INSTANT, International Nusantara* STratification ANd Transport program

1.  Introduction

The Indonesian seas provide a low-latitude pathway for the transfer of warm, low salinity Pacific waters into the Indian Ocean. This Indonesian Throughflow (ITF) plays an integral part in the global thermohaline circulation and climate phenomenon (see Sprintall et al., 2001 and Gordon, 2001 for recent overviews; Fig. 1), and the heat and freshwater carried by the ITF impacts the basin budgets of both the Pacific and the Indian Oceans (Bryden and Imawaki, 2001; Wijffels, 2001; Wajsowicz and Schneider, 2001). Within the internal Indonesian Seas, observations and models indicate that the primary ITF source is North Pacific thermocline water flowing through Makassar Strait (sill depth 650 m). Additional ITF contributions of lower thermocline water and deep water masses of direct South Pacific origin are derived through the eastern routes, via the Maluku and Halmahera Seas, with dense water overflow at the Lifamatola Passage (sill depth 1940 m). The ITF exits into the eastern Indian Ocean through the major passages along the Lesser Sunda Island chain: Ombai Strait (sill depth 3250 m), Lombok Strait (300 m), and Timor Passage (1890 m). The complex geography of the region, with multiple narrow constrictions connecting a series of large, deep basins, leads to a circuitous ITF pathway within the Indonesia seas. Enroute the Pacific inflow waters are modified before export to the Indian Ocean due to mixing, upwelling and air-sea fluxes.

In recent years a number of monitoring programs have measured aspects of the ITF from its Pacific source, through the internal seas, to the exit passages. The programs range from individual year-long mooring deployments in Makassar, Timor and Ombai Straits, a three-year shallow pressure gauge array (SPGA) in the exit passages, to decade long XBT transects within the Indonesian region and five full-depth hydrographic CTD/ADCP sections between Australia and Indonesia. A serious shortcoming in the recent measurement programs within the Indonesian region is the lack of temporal coherence: the data cover different time periods and depths in the different passages of the complex pathways toward the Indian Ocean. This has lead to ambiguity of the mean and variable nature of the ITF, and the transformation of the thermohaline and transport profiles within the interior seas. For this reason, and in keeping with the recommendations stemming from the full community deliberation (OCEANOBS October 1999 meeting in St. Raphael, see Imawaki et al., 1999; Workshop on Sustained Observations for Climate of the Indian Ocean, Perth, November 2000, see CLIVAR Exchange, 3:4, December 2000 and www.marine.csiro.au/conf/socio/socio.html), an international co-operative effort is planned to deploy multi-year moorings for direct Throughflow velocity, salinity and temperature measurements simultaneously in Makassar Strait, Lifamatola Passage, Lombok Strait, Ombai Strait and Timor Passage (Table 1). The program called INSTANT (International Nusantara STratification ANd Transport) involves 5 nations: Australia, France, Indonesia, Netherlands, and the USA. This collaborative proposal by Lamont-Doherty Earth Observatory (LDEO) and Scripps Institute of Oceanography (SIO) represents the U.S. effort of the international INSTANT experiment listed in Table 1. Letters of Intent from the other countries are included as appendices.

Figure 1.  Schematic of Indonesian Throughflow pathways (Gordon, 2001).  The solid arrows represent North Pacific thermocline water; the dashed arrows represent South Pacific lower thermocline water.  Transports in Sv (106m3s-1) are given in red.  The 10.5 Sv in italics is the sum of the flows through the Lesser Sunda passages.  ME is the Mindanao Eddy; HE is the Halmahera Eddy.  Superscript refers to reference source:  1: Makassar Strait transport in 1997 (Gordon et al., 1999); 2: Lombok Strait (Murray and Arief, 1988; Murray et al., 1989) from January 1985 to January 1986; 3: Timor Passage (between Timor and Australia) measured in March 1992 to April 1993 (Molcard et al., 1996); 4: Timor Passage, October 1987 and March 1988 (Cresswell et al., 1993); 5: Ombai Strait (north of Timor, between Timor and Alor Island) from December 1995 to December 1996 (Molcard et al., 2001); 6: between Java and Australia from 1983 to 1989 XBT data (Meyers et al., 1995; Meyers, 1996); 7: Upper 470 m of the South Equatorial Current in the eastern Indian Ocean in October 1987 (Quadfasel et al., 1996); 8: Average ITF within the South Equatorial Current defined by 5 WOCE WHP sections (Gordon et al., 1997).  The hollow arrow represents overflow of dense Pacific water across the Lifamatola Passage into the deep Banda Sea, which may amount to about 1 Sv (van Aken et al., 1988).  Inserts A-D show positions of INSTANT moorings. Insert A: 2 Makassar Strait Inflow moorings (U.S., red diamond) within Labani Channel. Insert C: Netherland’s mooring within the main channel of Lifamatola Passage (yellow triangle). Insert B, D: Sunda moorings in Ombai Strait, Lombok Strait, and Timor Passage (U.S., red diamonds; France, purple square; Australia, green circles).  The positions of the shallow pressure gauge array (U.S., green X).  The 100, 500, and 1000 m isobaths are shown in the inserts.

The collective merit for an internationally coherent program has become strongly apparent given the complexity and scope of the disparate nature of the present observations. INSTANT is designed to provide a time series of ITF transport and property fluxes, and their variability from intraseasonal to annual time scales, along the ITF pathway from the intake of Pacific water at Makassar Strait and Lifamatola Passage, to the Lesser Sunda exit channels into the Indian Ocean (Fig. 1). The international group of PIs and their respective responsibilities as part of the INSTANT program are listed in Table 1. All the PIs bring their own valuable scientific expertise to the project, and all have established good scientific track records of working in the Indonesian region on Throughflow problems, both individually and collaboratively as a team. Many of the supporting measurements of the INSTANT program, such as the multi-year time series from the Indonesian tide gauge and meteorological network, and the Australian XBT program, already exist and are ongoing. The combined data set from the international INSTANT mooring experiment and the supporting observational network, will leverage our individual measurements to understand the broad spectrum of variability in the ITF, and the role that regional oceanography plays in establishing the transfer function between the Pacific inflow and the outflow into the Indian Ocean.

 

Table 1.  Components of INSTANT

ITF Mooring Components:

  Makassar Strait

USA

Gordon, Ffield, Susanto

  Lombok Strait

USA

Sprintall

  Ombai Strait

USA

Sprintall

  Timor Passage

Australia

Wijffels

  Timor Passage

France

Molcard, Fieux

  Lifamatola Passage

Netherlands

van Aken

Supporting/Proxy Measurements:

  CTD and Underway ADCP

Indonesia

BPPT

  Lesser Sunda Pressure Gauges

USA

Sprintall

  XBT Network

Australia

Wijffels, Meyers

  Tide Gauges

Indonesia

Bakosurtanal (survey and mapping)

  Meteorological Network

Indonesia

Badan Meteorologi dan Geofisika

 ARGO Floats

Australia

Wijffels

Ship Facilities:

 

 

  Baruna Jaya

Indonesia

BPPT

 

The two Makassar moorings build upon the results of the December 1996 to July 1998 Arlindo Makassar moorings that were deployed in the Labani Channel, a 45 km wide, 2000 meter deep, constriction near 3°S (Fig. 1). Frictional considerations and local wind processes therefore lead to very little throughflow contribution over the very shallow coastal flats to the west of the moorings (Fig. 1). The INSTANT moorings in Makassar Strait will be deployed at the same sites as the Arlindo moorings, within the deep Labani channel.  The Lifamatola Passage mooring will resolve the lower thermocline and deeper water inflow into the Banda Sea. The Sunda observations in Lombok Strait, Ombai Strait, and Timor Passage, will capture the ITF as it exits from the interior seas into the Indian Ocean. Most moorings will be instrumented with upward-looking ADCPs to measure velocity of the surface layer. The Makassar and Lifamatola moorings will also have downward looking ADCPs to measure the deep, dense water overflows. The moorings will have a relatively dense suite of discrete velocity, temperature and salinity measurements designed to resolve the net ITF property transports.

The initial deployment of all moorings will be in August 2003, with re-deployment in February 2005, and recovery in July 2006. The resulting simultaneous, multi-year direct measurements in both the inflow and outflow ITF passages will allow us, for the first time, to unambiguously determine the magnitude of inter-ocean transport, its properties, and their evolution between the Pacific and the Indian Oceans. Finally, an important outcome of the program will be to recommend an efficient set of ongoing observations that may offer a proxy for an efficient long-term monitoring strategy of the ITF transport and property fluxes.

2.  Scientific Objectives of INSTANT

2.1  The Global Perspective: To determine the full depth velocity and property structure of the Throughflow and its associated heat and freshwater flux.

Large-scale observation based studies (including inverse solutions) reveal significant Pacific export of mass, heat and freshwater into the Indian Ocean (Piola and Gordon, 1984, 1986; Toole and Raymer, 1985; Wijffels et al., 1992; Toole and Warren, 1993; MacDonald, 1993; MacDonald and Wunsch, 1996; Ganachaud et al., 2000; Ganachaud and Wunsch, 2000). Uncertainty in the size of such a warm, fresh Throughflow is the dominant source of error in analyses of the basin-wide budgets of heat and freshwater for the Pacific (Wijffels et al., 2001) and Indian Oceans (Robbins and Toole, 1997). Models reveal dependence of Pacific and Indian Ocean SST and upper layer heat storage on the throughflow (Hirst and Godfrey, 1993; Verschell et al., 1995; Murtugudde et al., 1998). The Indian and Pacific Oceans would be very different if the ITF were zero (MacDonald, 1993; Schneider, 1998; Maes , 1998; Wajsowicz and Schneider, 2001). Oceanic heat and freshwater fluxes into the Indian Ocean - at the expense of the Pacific - affect atmosphere-ocean coupling with potential impacts on the ENSO and monsoon phenomena (Webster et al. 1998). SST anomalies which lead to a dipole mode of variability in the tropical Indian Ocean first appear in the vicinity of Lombok Strait in Indonesia (Saji et al., 1999).  The cold SST anomaly shifts westward along the Java and Sumatra Indian Ocean coast and varies in strength with the phase of ENSO (Susanto et al., 2001). The ITF source water (North Pacific vs. South Pacific) depends upon land geometry and the tropical Pacific wind fields (Nof, 1996; Wajsowicz, 1999; Morey et al., 1999; Cane and Molar, 2001). Observations show that the ITF is composed mostly of North Pacific water flowing through Makassar Strait (Fine, 1985; Ffield and Gordon, 1992; Gordon, 1995; Gordon and Fine, 1996), consistent with theory (Wajsowicz, 1996).

Makassar Strait Moorings:  The INSTANT moorings in Makassar Strait will be deployed at the same coordinates as the 1996-1998 Arlindo moorings. The Arlindo moorings provided current and temperature time series at various depths (Fig. 2) at two moorings within Labani Channel of Makassar Strait between December 1996 and June 1998 (Gordon et al., 1998; Gordon and Susanto, 1999; Gordon et al., 1999; Ffield et al., 2000; Susanto et al., 2000; Vranes et al. 2002). The 1997 average southward transport in Makassar Strait is 9.3±2.5 Sv.

The highest southward flow occurred at mid to lower thermocline depths (Gordon and Susanto, 1999; Gordon et al., 1999). The subsurface maximum occurs during times of large transport, from April to September 1997 and again in April 1998 to the end of the record in June 1998. An upward looking ADCP at 150 m on the Arlindo Makassar Strait moorings, provided a record of surface layer flow from 1 December 1996 to 1 March 1997. This time series shows increasing southward speeds with increasing depth, while the surface flow varied from near zero to northward. Attesting to the complex velocity profile, Waworuntu et al. (2001) - using TOPEX/POSEIDON altimeter data, and data from 2 pressure sensor/inverted echo sounders deployed in Makassar Strait as part of the Arlindo 1996-1998 array - find that a three layer model of the transport profile is required to explain the Makassar dynamics, with the middle layer (roughly 200-500 m) providing a significant contribution to the total transport. The subsurface velocity maximum through Makassar Strait results in a low transport-weighted temperature of 11.5°C, with an average salinity of 34.45 (Vranes et al., 2002). This is cooler and saltier than the ITF characteristics envisioned by Piola and Gordon (1984; ITF salinity of 33.6), and Toole and Warren (1993; ITF temperature of 24°C). In the southeast Indian Ocean, the transport-weighted temperature over the top 650 db across the IX1 XBT line is ~17.1°C, from the 15-year time series (Wijffels and Meyers, pers. comm.). Warming between the inflow at Makassar and the export into the Indian Ocean is likely a consequence of the impact of mixing, air-sea fluxes and upwelling within the internal Indonesian seas.

While some of the Makassar throughflow exits the Indonesian Sea within Lombok Channel (Murray and Arief, 1988), most turns eastward within the Flores Sea to enter the Banda Sea before entering the Indian Ocean via the deeper Ombai Strait and Timor Passage (Gordon and Fine, 1996). The Banda Sea receives small amounts of South Pacific lower thermocline and intermediate layers by way of the Maluku and Halmahera Seas (Ilahude and Gordon, 1996; Gordon and Fine, 1996; Ffield and Gordon, 1996; Hautala et al., 1996; Kashino et al., 1996; Cresswell and Luick, 2001; Luick and Cresswell, 2001). It may be that some Makassar transport flows northward east of Sulawesi to return to the Pacific Ocean, but it is unlikely that this is a significant amount: water mass analysis of the Arlindo data (Gordon and Fine, 1996) in that region does not indicate northward flow within the thermocline. Northward transport may occur in the surface layer, which may be composed in part of upwelled Makassar water in the Banda Sea. Annual average upwelling within the Banda Sea is about 0.75 Sv (Gordon and Susanto, 2001).  If half of the upwelling flows northward, less than 5% of the Makassar transport may return to the Pacific Ocean in the surface layer, not a significant amount. At depth, the Banda Sea is flooded with overflow of South Pacific waters across the 1940 m deep Lifamatola Passage (Jonker et al., 1986; van Aken et al., 1988). The ITF contribution through the Lifamatola Passage will be monitored by a full-depth mooring, contributed by the Netherlands.


 


Figure 2.  (a) Time series (above) of the average temperature (red) between 150 and 400 db at the MAK-1 mooring.  The SOI (green) and the Makassar Strait volume transport (blue dashed) are also shown.  The data are smoothed by 30 day running averages.  (b) Temperature time-section (right) constructed from 15 years of Makassar Strait and Flores Sea XBT profiles.  In the upper panel, the depth of the 22°C XBT isotherm (red) is shown with the SOI (black) highlighting the clear ENSO variability in the XBT temperature data.  The data are smoothed by a 1 year running average.



Lesser Sunda Moorings:  The flow in Lombok Strait is well documented, largely due to the Lombok Strait Experiment in 1985 (Murray and Arief, 1988; Murray et al., 1990). Flow through the 35 km wide Lombok Strait is restricted by a sill depth of ~ 300 m at the southern end. Murray and Arief (1988) deployed 5 moorings along two transects: one along the sill and another located north of the sill in ~1000 m of water. The moorings from the sill transect largely failed as the tidal currents were extreme, reaching amplitudes of 3m/s: too vigorous for the mooring technology of the day. Of the 2 moorings located north of the sill, all instruments on the eastern mooring functioned for the full deployment from January 1985 through January 1986, while the western mooring was only instrumented after June 1985. The tidal current amplitudes at the northern moorings were only 0.5-1 m/s (Murray et al., 1990), suggesting a more quiescent environment and thus a better site for our two proposed INSTANT moorings.

Flow at the northern Lombok Strait transect was predominantly southward, with strongest flows in the upper 100 m, and well-correlated down to 300 m (Murray and Arief, 1988). The upper 100 m carries one-half of the total transport through Lombok Strait (Murray and Arief, 1988). Flow at 800 m, below the sill depth, was weak and exhibited variations not obviously related to the shallower flow. From February through May, transport through Lombok Strait was only ~1 Sv southward due to discrete occasions of shallow northward flow, compared to the 4 Sv observed during the southeast monsoon from July through September (Murray and Arief, 1988). Similar periods of northward transport are inferred from changes in the cross-strait pressure gradient measured by the shallow pressure gauges deployed across Lombok Strait from 1996-1999 (Fig. 3; Hautala et al., 2001). The flow reversals observed in Lombok Strait are likely related to the passing of Kelvin waves, forced in the equatorial Indian Ocean (Sprintall et al., 1999; 2000).

 



 

 

Figure 3.  Time series of 25m geostrophic velocity and estimated upper layer (0-100m) transport for each of the straits.  The last panel shows two transport sums:  Lombok + Ombai (solid) and Lombok + Ombai + Timor (grey).

 


Recent ADCP cross-sections of the upper 250 m in Lombok Strait (Fig. 4) distinguish a more variable upper layer reaching to 100 m (with northward flow in December 1995), from a deeper layer that displayed constant southward velocities of 20-60 cm/s (Hautala et al., 2001). In all three survey years, the strongest outflow is found near Bali, on the western side of the strait. Time series of temperature and salinity, co-located at the pressure gauge sites in Lombok Strait, consistently find the Bali side to be cooler and saltier compared to Lombok. The distinct behavior of the upper/lower layers, and the quasi-permanent western intensification was also observed at the northern moorings during the southeast monsoon in the Lombok Strait Experiment (Murray and Arief, 1988; Arief, 1992). In fact the oft-cited average transport through Lombok Strait of 1.7 Sv (Murray and Arief, 1988) was determined as an average of current meter data from either side of the channel for 1985, although the western (Bali) mooring was only instrumented after June 1985. Their reported transport curve may therefore tend to underestimate the amplitude of the outflow pulse occurring near the end of the southeast monsoon, and consequently the annual average as well.

Ombai Strait lies between Alor and Timor Islands, and connects the Banda Sea with the Savu Sea (Fig. 1). The strait is 30 km wide, and a recent bathymetric survey by Molcard et al. (2001) suggests a sill depth of 3250 m. From a single mooring in Ombai Strait, Molcard et al. (2001) suggest a transport range between 4.3 Sv and 5.8 Sv, depending on the assumed cross-strait shear, with 0.9-1.2 Sv average transport in the upper 100 m. Although the deeper currents between 500-1220 m are weak, they still contribute to the transport and appear uncorrelated to surface variability, and thus must be independently measured.

The flow measured in Ombai Strait was directed westwards towards the Savu Sea (and Indian Ocean) during the entire mooring deployment, except at the beginning of the record in December 1995, when flow reversed in the upper 120 m towards the Banda Sea (Molcard et al., 2001). December 1995 coincides with the northward surface flow observed in Lombok Strait (Fig. 4), and the eastward surface flow across Sumba Strait at the western edge of the Savu Sea (Hautala et al., 2001). The reversed surface flow during December 1995 is most likely due to the seasonal eastward incursion of the South Java Current associated with the passage of the coastal Kelvin wave passing into the Savu Sea and on through Ombai Strait. Indeed, cross-correlations of the pressure gauge time series during December 1995, show the eastward propagation from Bali to Sumbawa (northern Sumba Strait) to the Ombai Strait gauges, with a speed of roughly 2.5 m/s, commensurate with linear wave theory (Chong et al., 2000). ADCP surveys across Ombai Strait in both March 1997 and March 1998 show the most persistent westward Throughflow in these passages is mainly confined to the southern two-thirds of the straits (Fig. 4). Thus at least two INSTANT moorings are required in Ombai Strait in order to capture the cross-passage variability.

Figure 4.  Tidally-corrected along-strait velocity and section-mean velocity profiles in Lombok Strait, Sumba Strait, Ombai Strait, and Timor Passage during December 1995 (top), March 1997 (middle), and March 1998 (lower panel), after Hautala et al. (2001).  In the velocity sections, the contour interval is 20 cm/s, and negative values are towards the Indian Ocean and shaded.  Error bars in the velocity profiles are light lines.  Arrows show the position of the moorings deployed in Timor Passage in 1992-93 (Molcard et al., 1996).

The Timor Passage is a long, narrow trench that lies between the south-east oriented coast of Timor and the northern edge of the wide, shallow Northwest Australian coastal shelf (Fig. 1). The eastern-most sill connecting the Timor Passage with the Banda Sea is 1250 m, shallower than the 1890 m western sill connecting the trench with the southeast Indian Ocean near Roti Island. Deployment of two year-long moorings across the western sill have been undertaken on two occasions by the French: a total westward transport of 4.5+/- 1.5 Sv from 120 -1050 m in August 1989 - September 1990 (Molcard et al., 1994); and 4.3 +/- 1 Sv westward from 0-1250 m in March 1992-April 1993 (Molcard et al., 1996). Based on an ADCP repeat survey of Timor Passage in December 1995 (Fig. 4), Hautala et al., (2001) suggest that the high-velocity surface core of the Throughflow may extend further south than the Molcard et al. (1996) extrapolation allowed for, that would lead to a much higher total transport through Timor Passage. Longer time series from four moorings near the western sill in Timor Passage, contributed by Australia and France as part of INSTANT, will more fully resolve the Timor Passage contribution to the Throughflow.

2.2  To resolve the annual, seasonal and intraseasonal characteristics of the ITF transport and property flux.

The recent measurements around the Indonesian archipelago revealed an unanticipated richness in the time-scales of Throughflow variability, from intraseasonal (40-60 days) to interannual (El Niño). The different time scales are evident in all the transport and property fluxes and are likely the result of remote forcing by both the Pacific and Indian Ocean winds, and local forcing within the regional Indonesian seas. Earlier estimates of the mean Throughflow were wide ranging (2-22 Sv; Godfrey, 1996) in part because of the lack of direct estimates, but also because of the real variation that can severely alias estimates of the mean if survey periods are not long enough. In fact, the recent estimates from the time series measurements suggest a Makassar Strait inflow transport of 9.3 Sv that is comparable to the transport sum of 10.5 Sv through the passages of the Lesser Sunda Island chain (Fig. 1). However, these mean ITF estimates should be interpreted with some caution as the time series were made at different times and at different periods of the ENSO phase: the Makassar moorings were in 1997-98; the Ombai mooring in 1996; Timor Passage in 1989-90 and 1992-93; and Lombok Strait in 1985.

Annual changes in the ITF arise from the directionally-varying winds associated with the regional monsoon system. During the northwest monsoon from December through March, surface waters flow into the Banda Sea (Wyrtki, 1987), with Ekman downwelling reaching a maximum in February (Gordon and Susanto, 2001). The flow from the Banda Sea towards the Indian Ocean is strongest from July to September, when flow is enhanced by the local Ekman response to the more intense southeast monsoon (Wyrtki, 1987). Observations from the 3-year pressure gauge data in the Lesser Sunda outflow passages (Fig. 3; Chong et al., 2000; Hautala et al., 2001), as well as year-long current meter data in Lombok Strait (Murray and Arief, 1988), Ombai Strait (Molcard et al., 2001) and Timor Passage (Molcard et al., 1994; 1996) all find maximum Throughflow occurs during the southeast monsoon period.

Remote wind forcing from the Indian Ocean is responsible for changes in the ITF via the generation of coastal Kelvin waves. Anomalous wind bursts in the equatorial Indian Ocean force an equatorial Kelvin wave that, upon impinging the west coast of Sumatra, results in poleward propagating coastal Kelvin waves. Westerly wind bursts force downwelling Kelvin waves semi-annually during the monsoon transitions, nominally in May and November, and intraseasonally in response to the Madden-Julian Oscillation. The poleward propagation of the downwelling coastal Kelvin wave raises the sea level along the Indonesian wave guide of Sumatra and the Lesser Sunda Islands (Clarke and Liu, 1993; 1994), causes maximum eastward flow of the semi-annually reversing South Java Current (Quadfasel and Cresswell, 1992), and may temporarily reverse the along-strait pressure gradient within the Lesser Sunda passages. In May 1997, a semi-annually forced Kelvin wave resulted in eastward flow of very fresh, warm water in the South Java Current (Sprintall et al., 1999), with subsequent northward flow observed up through Lombok Strait and past the mooring in Makassar Strait (Sprintall et al., 2000).

Remote winds in the Pacific Ocean drive large-scale circulation that impacts the ITF on low-frequency, interannual time scales. The weakening of Pacific tradewinds that occurs during El Niño, reduces the pressure gradient between the Pacific and the Indian Ocean thought to be the driving force for the Throughflow on long time scales (Wyrtki, 1987). The changes are of the sense that transport is smaller (larger), and the thermocline is shallower (deeper) during El Niño (La Niña) time periods (Kindle et al., 1989; Meyers, 1996; Bray et al., 1996; Fieux et al., 1996; Gordon and Fine, 1996; Potemra et al., 1997). The high resolution POP model yields a 12 Sv annual average during La Niña and 4 Sv average during El Niño (Gordon and McClean, 1999). The Arlindo mooring observations within Makassar Strait, which span the entire cycle of the strong 1997/1998 El Niño, find a correlation (r = 0.73) between Makassar transport and ENSO (Gordon et al., 1999; though the time series is far too short to say this with assurance). During the El Niño months of December 1997 to February 1998, the transport average is 5.1 Sv with a corresponding internal energy ‘heat’ transport of 0.39 pW, while during the La Niña months of December 1996 to February 1997, the average is 12.5 Sv, a 2.5-fold difference, with a 0.63 pW heat transport. In the Lesser Sunda exit passages, the transport variations estimated from the shallow pressure gauges suggest diminished flow through Lombok and Ombai Straits during the ENSO years 1997-98, with increased flow through Timor Passage (Fig. 3, Hautala et al., 2001). This is consistent with a 20-year run of the POCM model, in which there is stronger flow through Timor Passage and weaker flow through Ombai Strait during ENSO events (Potemra et al., 2002), although the magnitude of the transport changes through the two model straits do not fully compensate each other. This change in the ratio of transport carried through each of the Indonesian passages has been hinted at by the recent measurements, although again, the lack of simultaneous direct measurements in the passages makes interpretation difficult. For example, water mass analysis suggests that during La Niña time periods, when Makassar Strait Throughflow is maximum, it appears that additional, albeit small, amounts of North Pacific thermocline water may enter the Banda Sea via the Lifamatola Passage from the Maluku Sea (Gordon and Fine, 1996).

In order to resolve the time scales of variability in the ITF transport and property flux, it is necessary for simultaneous direct measurement within the major Indonesian Throughflow passages, for a minimum period of three years. The characteristics of the annual, semi-annual and intraseasonal flow during the deployment period will be directly captured. While a 3-year time series alone is not sufficient to comprehensively resolve the interannual signal, combination with the earlier direct measurements in Makassar Strait (Gordon et al., 1999), Lifamatola Passage (van Aken et al., 1988), Lombok Strait (Murray and Arief, 1988), Ombai Strait (Molcard et al., 2001), and Timor Passage (Molcard et al., 1994, 1996), can help put the observational period into perspective. Further, with simultaneous monitoring the propagation of features such as coastal Kelvin waves along the wave guide of the Lesser Sundas and up into Makassar Strait and the internal seas should be readily apparent. Finally, the transport partitioning through the various passages can be explicitly determined.

2.3  To investigate the storage and modification of the ITF waters within the internal Indonesian seas, from their Pacific source characteristics to the Indonesian Throughflow water exported into the Indian Ocean.

The different monitoring programs within the Indonesian region indicate large differences in peak transport timing between the inflow and outflow straits. For example, while Gordon et al. (1999) found maximum transport through Makassar Strait during the northwest monsoon of 1996-1997, flow observed during the SPGA leaving the interior seas for the Indian Ocean was minimum during the same period (Hautala et al., 2001). The phasing differences are most likely due to storage in the Banda Sea. Assuming a two-layer ocean, Gordon and Susanto (2001) find Banda Sea surface water divergence, estimated from satellite altimetric data, correlates reasonably well with a 3 month lag of the export of upper ocean water through the Lesser Sunda Island exit passages, as estimated by the SPGA. It appears that the Banda Sea acts as a reservoir, filling up and deepening the thermocline during the northwest monsoon. During the more intense southeast monsoon, Ekman flow in the Banda Sea combined with the lower sea level off the south coast of the Lesser Sunda Islands, are more conducive to drawing waters into the Indian Ocean (Wyrtki, 1987). The storage of mass and heat within the internal Indonesian seas will dramatically affect the interpretation of the measurements of the Throughflow if made over short time scales.

The Banda Sea is also the primary site for conversion of the ITF profile from Pacific into the distinct Indonesian Sea stratification, which is then exported into the Indian Ocean (Ffield and Gordon, 1992). A series of CTD stations from the Arlindo 1993-94 cruises shows the attenuation and modification of the temperature and salinity profiles from the North Pacific salinity-maximum (100-150 m) and salinity-minimum core layers (250-350 m) with distance from the Makassar Strait (Ilahude and Gordon, 1996). The temperature and salinity stratification as well as the SST, are significantly altered by the strong air-sea fluxes, seasonal wind-induced upwelling and large tidal forces within the Banda Sea (Ffield and Gordon, 1992; 1996; Hautala et al., 2001). Besides the vertical mixing, the attenuation of the salinity minimum is also due to the isopycnal injection of the more saline water from the South Pacific that enters the Banda Sea through Lifamatola Passage (Gordon and Fine, 1996; Hautala et al., 1996).

Hence it appears that both the composition and magnitude of the stored waters within the Banda Sea could have a significant impact on the Indian Ocean heat, freshwater and mass budgets. For example, Hautala et al. (2001) suggest that an imposed transport imbalance of 5 Sv, consistent with that determined by Gordon and Susanto (2001) on seasonal time scales, affects the temperature at thermocline depths by about 5°C. Convergences and divergences of this magnitude can have a substantial impact on thermocline stratification, and thus may affect SST by changing the temperature of the water being entrained into the mixed layer.  Because of the possibility of different sampling biases in the previous measurements of the inflow and outflow straits, and the relatively short time series, we cannot as yet unambiguously determine transport imbalances into and out of the internal seas. Part of the apparent differences in transport timing between the inflow and outflow straits may be due to the nature of the measurements: the Makassar instruments were largely below 200 m, whereas the SPGA in the outflow straits samples the upper Throughflow. Another cause for the difference in peak-to-peak timing between transport through Makassar Strait and the exit passages, may be the contribution to the ITF through the seas and passages directly north of the Banda Sea, the so-called “eastern” route for the ITF. The significant phase differences between the inflow and outflow straits need to be tested with simultaneous direct measurements of both surface and subsurface velocity and property measurements in all the major Throughflow passages, over at least a 2-3 year time period, as proposed here in the INSTANT program. Combined with remotely-sensed SST, altimeter and scatterometer data, this should enable us to address the issues of implied local storage within the Indonesian seas, along with the changes in the water mass characteristics in the internal basins that may occur through mixing.

2.4 Contribute to the design of a cost-effective, long-term monitoring strategy for the ITF

Techniques for developing proxy-ITF monitoring are ultimately needed in order to develop cost-effective, long-term ITF transport information. All measurements within the Indonesian passages require assumptions about the cross-passage vertical structure to link their data to the full passage velocity and transport information required to determine the ITF transport of mass and heat. This is true whether the observations are direct current meter measurements from single mooring deployments, or indirect measurements that need geostrophic assumptions. In addition, many observational programs can be logistically difficult and expensive to maintain over time-scales long enough to be important to climate variability. The challenge is to develop a relatively inexpensive but effective network for monitoring the ITF over long time-scales, that adequately captures the dominant time scales of throughflow dynamics, and in turn, improves our ability to understand regional climate variability.

Given the large amount of subsurface temperature data collected in the region along the XBT transects, as well as from the discrete moorings in Makassar Strait, one feasible approach for an ITF proxy-indicator comes from linking changes in the thermal structure to ITF transport. The Arlindo Makassar Strait time series reveals that the ITF transport is linked to thermocline depth (Fig. 2a): transport is smaller and thermocline shallower during El Niño (Bray et al., 1996, 1997; Meyers, 1996; Ffield et al., 2000). The correlation between variability in the average thermocline temperature to variability in the southward Makassar transport is r=0.67 (Ffield et al., 2000). Using nearly 15 years of XBT data, Ffield et al. (2000; Fig. 2b) show that the Makassar upper thermocline temperature is highly correlated with ENSO: 0.77 for the Southern Oscillation Index; -0.80 for NINO3 SST anomaly, and -0.82 for NINO4 SST anomaly. The correlations increase when the ENSO time series are lagged a month or so. Thus the Makassar temperature field - when coupled with the throughflow - transmits equatorial Pacific El Niño and La Niña temperature fluctuations into the Indian Ocean.

Another feasible technique for proxy ITF monitoring comes from linking the changes in ITF transport to the cross-strait changes in observed or remotely sensed sea level, or (equivalently) pressure gauge data. The technique has been successfully employed from observed shallow pressure variations in the SPGA, deployed in the exit passages of Indonesia from 1996-99 (Fig. 3; Chong et al., 2000; Hautala et al., 2001; Potemra et al., 2002). The transport fluctuations through the straits are calculated assuming geostrophy (confirmed by Potemra et al., 2002) to relate the cross-strait pressure gradient to velocity (Hautala et al., 2001). Using sea level to estimate the transport has two real advantages: the geostrophic estimate is an integral across each passage and the measurement is simple and relatively inexpensive to make. The disadvantage is that the vertical position of the gauges relative to the geoid is unknown, so only sea level fluctuations are determined. However, Hautala et al. (2001) used the information from concurrent repeat ship-board ADCP surveys of 1-2 days duration across the same passages as the pressure sensors (Fig. 4) to calibrate the sea level fluctuations, by scaling the inferred surface flow to the tidally-corrected ADCP measured surface flow. This yields the absolute shallow velocity and transport estimates shown in Fig. 3.  For the INSTANT program, we propose to redeploy the SPGA simultaneously with the mooring deployments in Lombok Strait, Ombai Strait and Timor Passage. Instead of using the repeat ship-board ADCP surveys to calibrate the sealevel fluctuations, in INSTANT we will use the 3-year time series from the concurrent mooring data. This approach has numerous advantages: the tidal signal, which is the largest source of error (10-20 cm/s) for the absolute velocity measurements (Hautala et al., 2001), can be more accurately resolved; the mean estimate and variability of the throughflow is available over at least an annual time-scale; the mooring data give valuable information about the variability of the cross-passage vertical structure over longer time-periods and the deeper flow below the upper 250 m and so not  measured by the ADCP.

Satellite altimetric measurements of sea surface height may also provide an effective proxy approach for determining ITF transport (see Potemra, 1999). Susanto et al. (1999) find a high correlation (r=0.7) between western Pacific altimetric sea surface height and Makassar Strait transport. Susanto et al. (1999) found their mode 1 (68% of the total variance), from a complex singular value decomposition of the altimetric sea surface height within the internal Indonesian seas, is well correlated with the NINO3 index. Mode 2 (15% of the total variance) is associated with the annual (monsoon) frequency. Similarly, Waworuntu (1999) shows a strong signal in the altimetric sea surface height signal of the central Banda Sea during the 1997-98 ENSO event, that corresponds to the same trends in Throughflow transport observed in Makassar Strait by Gordon et al. (1999): low (high) sea level corresponding to maximum (minimum) transport. Recently Gordon and Susanto (2001) determined that satellite altimetry may serve as an effective means for monitoring the Banda Sea surface layer divergences that may be a good indicator of storage in the internal seas. Finally, continuous, high-quality tide gauge data along the islands in the internal seas as provided by the Bakosurtanal (see Table 1 for supporting measurements) would be particularly useful for determining the dominant tidal frequencies needed for ground truthing the satellite data. The direct measurements of the ITF by current meter and ADCP moorings, such as proposed for INSTANT, will be necessary to help mold and ground-truth the required algorithm needed to convert the proxy data to ITF transport information. A two-three year coherent process study throughout the Indonesian region is needed to clearly identify the ITF proxies necessary to provide sustained observations.

3.  INSTANT Mooring Array (2003-2006)

Attempts to measure the ITF have been piecemeal: individual channels measured at different times. Thus we have no unambiguous picture of the mean and variable nature of the ITF. The proposed INSTANT program has the central goal of alleviating this deficiency, and to establish proxy indicators of the ITF which may be suitable for long term monitoring. Here we describe the design and instrumentation of the Makassar, Lombok and Ombai moorings as part of the U.S. involvement in the INSTANT program (Table 1).

3.1  Makassar Strait

We propose to install two moorings at the same positions as the Arlindo moorings of 1996-1998 (Fig. 1), with temperature, conductivity, and pressure sensors, current meters, and up- and down-looking  ADCPs (Table 2). The 500-650 m deep Dewakang Sill in the southern Makassar Strait is a heavily fished region, too broad with a labyrinth of channels to realistically measure the Throughflow. The extension of the Arlindo MAK time series by INSTANT Makassar time series will provide improved resolution of the interannual, seasonal and intraseasonal signals within the throughflow and their relationship to the temperature and salinity stratification.  In addition, essential information on the currents of the upper 200 m, not fully resolved by MAK-1 and MAK-2, will be obtained by the paired ADCPs.

The upward looking ADCPs (RDI Long Ranger 75 KHz)  will resolve the surface layer and upper thermocline.  We suspect that the currents in the upper 100 meters vary seasonally.  The downward looking ADCPs (WH300) will resolve the middle and lower thermocline currents.  The paired ADCPs are positioned on the mooring to profile what is hypothesized to be the maximum in throughflow current while keeping the bulk of the mooring flotation and associated drag in the lower speed regime of the lower thermocline.  Together, they will provide velocity profiles of the water column lying approximately between 50 and 400 meters during all phases of the tidally-induced mooring motion, minimizing ambiguities due to mooring blowover in this portion of the water column.  Current meters at 200 (MAK West) and 750 meters correspond to instrument depths of MAK-1, 2. The current meter placed at 1500 m on MAK West will allow for investigation of a phenemenon observed in the MAK-1,2 times series:  the 750 m and 1500 m along channel velocities, while averaging near zero, display out of phase events with speeds of up to 20 cm/sec.  The ADCPs and current meters will sample at 30 minute intervals.

 

Table 2.  INSTANT Mooring Array (US moorings shown in bold)

 

sill depth (m)

width (km)

 

mooring position (°S; °E)

bottom depth (m)

 

ADCP (+T)

 

currents (C),  temperature (T),  salinity (S),  pressure (p)

Makassar Strait

 

650

45

02°51.7'; 118°27.5'

(MAK West)

2137

300 (up)

310 (dn)

T-p: 45, 60, 80, 100, 130, 150, 170, 205, 225, 245, 295, 332, 366, 434, 468;  T-S-p: 115, 265;  C-T-p: 40, 200, 400;  C: 750, 1500

 

 

 

02°51.2'; 118°37.7'

(MAK East)

1611

300 (up)

310 (dn)

T-p: 295;  T-S-p: 115;  C-T-p: 400, 750

Lifamatola

1940

90

01°49’; 126°57’

1938

500 (up)

C-T-p: 600,800,1000

Passage

 

 

 

 

1500 (dn)

 

Timor

1890

80

11°09’; 122°37’

  800

 

T: 200,300;  C-T-S-p: 150, 250,400;  C-T-p: 800

Passage

 

 

11°15’; 122°42’

1850

250 (up)

C-T-S-p: 150, 400, 700;  C-T-p: 1000, 1500

 

 

 

11°25’; 122°47’

1500

250 (up)

T: 200,300;  C-T-S-p: 150, 400, 700;  C-T-p:  1000

 

 

 

11°35’; 123°52’

  700

 

T: 200,300;  C-T-S-p: 150, 400;  C-T-p:650

Ombai Strait

3250

30

8° 24’; 125°08’

1500

250 (up)

T: 200,300,500;  C-T-S-p: 150, 400;  C-T-p: 700, 1000

 

 

 

8° 33’; 125°09’

3200

250 (up)

T: 200,300,500;  C-T-S-p: 150, 400;  C-T-p: 700, 1000, 1500, 3000

Lombok

300

35

8° 24’; 115°44’

  800

200 (up)

T: 175,250,300;  C-T-S-p: 150, 450

Strait

 

 

8° 24’; 115°54’

1200

200 (up)

T: 175,250,300;  C-T-S-p: 150, 450;  C-T-p: 800

 

Fifteen Seabird SBE39 temperature-pressure recorders are distributed on MAK West to resolve the temperature stratification with a 5 minute sampling rate from the near surface through the thermocline.  Combined with the  temperature and pressure sensors on the current meters and the moored SBE37 Microcat CTDs (with sampling rates between 7.5 and 30 minutes) a complete thermocline temperature field will be resolved.  Strong semi-diurnal tidal currents will push the mooring over each tidal cycle, so all the sensors will profile the temperature stratification four times daily (Ffield et al. 2000).  A nearly continuous, high resolution, temperature-time section will be obtained.  The twice-daily mooring blowover due to the tides essentially converts the moored sensors into independent temperature profiling devices.  With the dense suite of temperature measurements planned here, tidal aliasing is minimized, and the Makassar Strait temperature stratification will be well resolved in the thermocline.

3.2  Lesser Sunda Island chain

The INSTANT mooring array includes 8 moorings within the three major straits of the Lesser Sunda Islands that control outflow into the Indian Ocean through the Indonesian archipelago: Lombok Strait (USA), Ombai Strait (USA) and Timor Passage (Australia, France; Fig. 1, Table 1, Table 2).

The location and number of moorings in Lombok and Ombai Straits, and the proposed depth of the instruments are designed to capture the full-depth velocity, temperature and salinity structure of the ITF outflow following the rationale outlined in the scientific objectives above. They are based on all the available information from the previous measurement programs conducted within the Lesser Sunda export passages. Thus, two moorings are proposed in both Lombok and Ombai Straits: on the eastern and northern sides, respectively  to capture the expected reversals due to the impact of the South Java Current, and along the western and southern sides to resolve the consistent intensification of ITF outflow (Fig. 4; Hautala et al., 2001). In Lombok Strait, the moorings will be located close to the northern moorings of Murray and Arief (1988). It may appear that measuring the ITF across Sumba, Dao and Savu Straits at the western edge of the Savu Sea, would be better than the proposed measurement at Ombai Strait on the eastern edge of the Savu Sea (Fig. 1). However, for this proposal, deploying a mooring array across these three wider channels is unrealistic and prohibitively expensive. In addition, the results from the inferred SPGA transport time series, suggests that the inflow from the eastern Ombai Strait and outflow through the three western straits should be balanced, at least on annual time scales (Hautala et al., 2001). As such, the 30 km wide Ombai Strait offers a convenient choke-point to measure the ITF velocity and property profiles as it directly exits from the Banda Sea.

Evidence from the recent measurements within the outflow straits suggest that, although the velocity does vary with depth, most of the transport seems to occur in a surface layer of only a few hundred meters thick. To capture this surface flow, all U.S. moorings will be outfitted with an upward-looking ADCP at 250 m in Ombai, and 200 m in Lombok. The ADCP will be set at surface-tracking mode so that the measurements can be corrected for the anticipated vertical motion due to tidal blow-over. The ADCPs have proven very successful in resolving the highly-energetic surface flow through both Ombai Strait and Timor Passage (Molcard et al., 1996; 2001).  WMI acoustic current meters in the surface layer at 150 m will provide backup for the ADCP. Discrete VMCM measurements will be made at intermediate depths and with RCM-8s at depth (Table 2), to adequately resolve the vertical velocity profile, and the lower thermocline contribution to ITF transport suggested by recent water mass analysis (Fieux et al., 1996; Wijffels et al., 2001).

Vertical variation in velocity is accompanied by variation in temperature and salinity. All  current meters and the ADCPs will be equipped with thermistors to resolve the temperature stratification. Additional temperature loggers will resolve the upper thermocline (Table 2). In addition, SBE-39 Microcat CTDs and  SBE-04 conductivity sensors will be distributed on the current meters. Expected strong tidal motion will be corrected for using co-located pressure sensors. Sampling rate for the ADCP, current meters, temperature and conductivity sensors will be 30 minutes. Thus properties are measured on the same space and time scales as the currents, so that net property transports can be estimated at all resolved frequencies.

3.3  Mooring Deployment Timetable and Hull-Mounted ADCP and CTD Station Plan

The initial deployment of all the INSTANT moorings will be in August 2003, recovery and re-deployment in February 2005, and final recovery in July 2006. The cruises for the Makassar and Sunda moorings will be coordinated so that ship-time and port-calls are optimized (see budget justification). The deployments and recoveries will be carried out from one of  the Indonesian BPPT Baruna Jaya ships (Table 1). The Baruna Jaya ships are modern research vessels, with ship-board ADCPs and capabilities for undertaking a hydrographic CTD survey of the region. The aft deck is considerable and has a large A-frame suitable for mooring work, indeed the PIs have previous experience in mooring deployment and recovery aboard the Baruna Jaya IV: Sprintall in the South Java Current (see Sprintall et al., 1999) and Gordon in Makassar Strait (see Gordon and Susanto, 1999; Gordon et al., 1999). The vessel used for recovery will be equipped with suitable gear for dragging for the mooring (trawl wire and winch) in the event both mooring releases fail.

The cruises include ship days for CTD/ADCP surveys of the channels. BPPT will carry out the CTD stations with their equipment and personnel. Concurrent shipboard ADCP and CTD surveys will be taken around the Makassar and Sunda moorings during each deployment/recovery cruise to provide important information on top-to-bottom water mass structure, and for calibration purposes. During the three mooring cruises we will steam back and forth across the entire channel width operating the Baruna Jaya hull ADCP over 1 to 3 diurnal tidal cycles. In addition, BPPT will obtain CTD stations to assess the along channel flow and stratification as a function of distance from the channel side walls.  Such data may be used as a basis for a correction factor for the transport calculations. This will extend the stratification data to further build the 1994-96 Arlindo CTD time series of Makassar Strait and the Indonesian internal seas, and of the French JADE hydrographic and the SPGA data in the Lesser Sunda export channels.

LDEO and SIO personnel will carry out the mooring preparation, deployments, and recoveries, with additional support via a subcontract to the CSIRO Marine Research (CMR) mooring group, supervised by Dr. Susan Wijffels. The mooring designs (see Budget Justification for details) are a balance of maximizing data return, while minimizing blow-over, costs, and complexity of deployment in consideration to the deck space and capability of the Indonesian research vessel to be used in the program. The CMR subcontract will provide all the mooring hardware, the winches necessary to deploy the moorings, the bulk of the floatation, as well as the lease of most of the instrumentation. The purchase of additional current meters with temperature and conductivity sensors, ARGOS beacons and acoustic releases are necessary to complete the mooring array. The CMR subcontract for the mooring hardware and assembly makes the proposed work not only cost-effective as few additional instruments need to be purchased, but the CMR group brings their own experience to the project. The subcontract includes participation by 2 CMR mooring specialists, to assist with the cruises, who have previously worked on the Baruna Jaya IV during the Australian-Indonesian ASEAN mooring program (see Luick and Cresswell, 2001; Cresswell and Luick, 2001). In addition, Wijffels as the CMR supervisor, will bring her own valuable scientific expertise to the project, in the collaborative analysis of the acquired mooring data at no extra cost to this proposal. Wijffels has a distinguished scientific record of working in the Indo-Australian region on Throughflow problems (see attached resume): she is the chief PI responsible for many Australian projects in the region (see Table 1), such as the Indonesian seas XBT program, and the ARGO profiling float array in the Indian Ocean, of which she is the Australian representative on the International ARGO Scientific Steering Committee.

3.4  Proxy and Supporting Measurements in the Indonesian Region

An important practical outcome of the ITF program will be to recommend an efficient set of ongoing observations that adequately and cost-effectively monitor the mass and property fluxes between the tropical Pacific and Indian Oceans. As part of this proposal, the regional Indonesian seas XBT, tide gauge, and satellite altimetric data will be explored for their potential as proxies of Throughflow transport (Table 1). These data have the advantage of including both short and long period variations of the Throughflow properties, as well as being part of ongoing monitoring programs.  If the Makassar thermocline temperature to transport to ENSO correlation found during the Arlindo moorings can be confirmed over longer time periods with additional mooring data, then the proxy data can be used to estimate the Makassar thermocline temperature, volume transport, and internal energy transport from the mid-1980’s to present.

We are also proposing to redeploy the shallow pressure gauge array (SPGA) in the three outflow straits to thoroughly test its viability for inexpensive and sustained monitoring of ITF variability. (The pressure gauges, which rely on geostrophic assumptions for determining flow, are probably inappropriate for monitoring Makassar Strait given its proximity to the equator.) The pressure gauges represent the least expensive component of the proposed fieldwork (< $40K). Pressure is measured by a Paroscientific quartz sensor, and their high precision (0.3 mbar) provides significantly increased resolution of sea level over traditional stilling-well tide gauges. The largest error comes from instrument drift, which at 0.3 mbar per year is much smaller than the dynamic signal of ~10 mbar found in the outflow straits (Hautala et al., 2001). The instruments are internally recording, and easily deployed by divers as they are held in the central column of a single-wheel anchor in a few meters water depth. In fact five of the six anchors are already in place in the three outflow straits from the 1996-99 SPGA deployment. No surface marker is used to locate the anchor, in order to reduce the problem of vandalism. Rather, in the initial deployment, an acoustic pinger will be attached to the anchor to make finding them in later cruises less of a challenge.

4.  Data Sharing and Outreach

Data Sharing: The INSTANT data will be shared among the participants as rapidly as preliminary numbers can be produced (anticipated within 3 months of mooring recovery cruises). A web site will be maintained at LDEO on which to post preliminary data. While most of the group interaction will be achieved through e-mail, workshops and presentations to the community are planned in conjunction with the American Geophysics Union and European Geophysical Society meetings. In December 2002, a planning meeting will be held during the AGU Fall meeting. In February 2006, a workshop is planned at the AGU Ocean Sciences meeting to discuss preliminary results from the first 18 month deployment. In April 2007, the INSTANT full 3 year results will be presented to the community at the European Geophysical Society meeting; a special session on the Indonesian Throughflow will be requested. Immediately prior to the EGS meeting a workshop of all INSTANT investigators will be held in Nice. The full INSTANT data set will be made available to the community 2 years after the moorings final recovery.

Outreach: It is important that we convey to the public the excitement and importance of ocean sciences in the remote, exotic reaches of Indonesia. We plan the following activities. Susanto will give ITF lectures to Indonesian schools and governmental agencies while he is in Indonesia participating in the INSTANT cruises. Gordon will incorporate ITF information in his Columbia University undergraduate course on the Earth's Climate, offering independent research projects. Sprintall will offer research opportunities to an undergraduate student as part of the Scripps undergraduate research fellowship (SURF) program. ITF exhibits are planned for Lamont (Ffield) and Scripps (Sprintall) open houses.

5.  In Conclusion:

INSTANT is an ambitious collaboration of 5 nations (Australia, France, Indonesia, Netherlands and the USA), which is designed to provide a time series of Indonesian throughflow transport and property fluxes, and their variability from intraseasonal to annual time scales, from the intake of Pacific water at Makassar Strait and Lifamatola Passage, to the Lesser Sunda exit channels into the Indian Ocean. The collective merit for an internationally coherent program has become strongly apparent given the complexity and scope of the disparate nature of the present observations, and the importance of arriving at a meaningful ITF. The international group of PIs bring their own valuable scientific expertise to the project, and all have established good scientific track records of working in the Indonesian region on Throughflow problems, both individually and collaboratively as a team.

Working in remote areas of the world is always a challenge. Recent progress in establishing official joint programs between Indonesia and the U.S. should be noted: both SIO and LDEO have Memorandums of Understanding with the Agency for Assessment and Application of Technology (BPPT), the primary agency responsible for management of the research vessels and oceanographic development within Indonesia. This progress, plus the PIs previous experience in conducting oceanographic research within Indonesian waters, augers well for continuing cooperation to do the proposed work. As the correspondence in Appendix I indicates, Dr. Indroyono Soesilo, the Chairman of the Marine Research and Fisheries Agency (within the newly organized Indonesian government Department of Maritime Affairs and Fisheries) is enthusiastic about the possibility of the proposed work. The involvement of Indonesian scientists will benefit both the proposed program implementation and the development of oceanographic research in Indonesia.



* Nusantara – Archipelago of Indonesia