(1) Meeting format. A limited number of invited talks in plenary bracketed each day. Science and technical working groups were organized with non-overlapping meeting times so that all could participate in at least one of each. Plenary sessions covered brief descriptions of data centers/products and large models and summaries of working group discussions.

RECOMMENDATIONS: The invited talks were very well-received. Exchange of ideas between groups which usually do not communicate (for instance, observers and modelers, observers of one type and those of another) was greatly facilitated by the content of the talks.

The division into science and technical groups was useful where specific data products needed to be discussed separately from the science discussions. However the model/model/data comparison technical group overlapped heavily with the science groups, probably because the topics for the science working groups were not geared well enough to the topics of interest to the upper ocean/ TOPEX/modeling groups. More emphasis in the science groups on these topics would have been useful.

Plenary session discussions were cancelled for most of the week to allow more time for the working groups and to allow "free time" for smaller group discussions and work. The final summary plenary session was well attended. Slightly more time in plenary could have been included as part of the invited talk sessions in order to summarize very briefly the daily progress of the working groups.

Many participants would have preferred more "free" working time. A suggestion is that much more time be allowed in the afternoon for this, moving the afternoon speaker to the evening.

Although it was indicated in the meeting announcement that attendance at the workshop would be limited, in practice we did not turn anyone away. This meant that our attendance was larger than original expectations, and so the working group structures became somewhat cumbersome with a decrease in the ability to be very flexible. However, it also meant that almost every major WOCE program for the Pacific as well as the agencies funding WOCE programs were well-represented at the workshop. The benefits of the presence of so many scientists engaged in Pacific research far outweighed the difficulties with the working group structures.

(2) Working group organization. Some groups established schedules for presentations in advance of the workshop. Others just requested that people make presentations on the spur of the moment. The latter worked well in the smallest groups, but was unsatisfactory in the larger working groups.

RECOMMENDATIONS: Future workshop working groups should make firm schedules of presentations several weeks in advance of the workshop, allowing considerable time for discussion. The workshop agenda and presentation schedules should allow time for new groups to be formed mid-workshop.

Working group leaders should be appointed well in advance.

Another possibility would be poster presentations in each working group at the beginning of the week. Posters could remain up for the whole meeting.

(3) Working group topics and size. Daily evaluations of each of the working groups were made by the organizing committee and working group leaders, leading to formation of new working groups and reduction in number of meetings of some groups during the week. Science topics of most interest to the largest scale modeling and observations (e.g. GCM's, TOPEX/Poseidon, drifters, broad-scale XBT's, etc) were unfortunately scheduled in a very large working group, which disintegrated. This left a void in science topics for a number of people, which was filled by several smaller group topics which formed mid-week and by the afternoon model/model/data comparison technical working group. Regardless of topic, the groups which worked well (lively discussion, new ideas, projects to be carried forward) were composed of about 20 or fewer people. Groups which consisted of 40 or more people were singularly unable to generate discussion or spontaneous presentations.

RECOMMENDATIONS: No working group should be larger than 20 people. Larger gatherings should be structured as firmly focused and planned plenary sessions.

Working groups which hope to have vigorous discussion should be seated in a conference style configuration.

Daily evaluation was necessary and highly recommended for future workshops.

(4) Computer network. Details and recommendations about the computer network can be found below in section II.4. Foreknowledge of the computational capabilities encouraged assembly of data and model products prior to the meeting. The set of workstations, peripherals and hubs installed in a central area, open 24 hours, was where computing actually occurred. Working groups did not use the network during their actual meetings.

RECOMMENDATION: The network and data/model product assemblage was clearly beneficial as it focused data assembly prior to the workshop and permitted data exchange during the workshop. It is recommended for future workshops. Detailed recommendations are listed below.

(5) Workshop venue and costs. The workshop was held at a resort hotel. This had the useful consequence of minimizing participant scatter. It was also somewhat too expensive for many. The meeting rooms were nominally free of charge as is customary with a hotel. However, the network installation costs were considerably larger than they would have been at a university. Guards were hired to safeguard the equipment.

There was no registration fee for the workshop. If the meeting had been held in a conference facility, then a nominal registration fee to pay for the rooms would have been a reasonable approach.

The total cost of the workshop to the U.S. WOCE Office was approximately $15,000. The computer manager's salary is not included in this total; he spent approximately 2 dedicated months preparing for and carrying out the meeting.

RECOMMENDATIONS: This venue worked well for communication and discussions, which continued visibly late into the evening on the hotel site. An academic conference center or retreat would be equally viable.

An initial set of working groups was selected in advance of the meeting. Some working groups established firm agendas prior to the meeting, with time scheduled for presentations from each member and much time for discussion. Other working groups did not make such a schedule, and instead asked for contributions from the floor during each meeting of the working groups. Only the smallest working groups (15 or less) worked well with this latter strategy.

The organizing committee and working group chairmen met each day over breakfast to discuss the progress of the workshop. By the second day it was clear that the largest groups should be reorganized. Several new topics had also emerged from the various groups and invited talks. New working groups were established, some for just one day, for specialized discussions. These groups worked very well.

Many groups brought sophisticated data and model displays to the workshop. In order to publicize what was available, there was a session on the first day for each DAC and SAC to present the data sets and data tools which they have made available (Appendix A.1). A similar session for presenting brief overviews of models and types of model results was inserted in the second day (Appendix A.2).

II.4. Computer network

The network consisted of an ethernet with connectors in 9 conference rooms, linked by a 128Kbit ISDN line to the internet. Internet service was contracted locally through cerfnet. Hubs in each conference room supported 8 to 16 hookups. Most of the 6 unix workstations, 1 Macintosh, 1 X-terminal, hubs, hard disks, cdroms, DAT drive, Exabyte drive and printers (2 color, 2 black and white) were loaned on a voluntary basis by scientists; the exceptions were a SGI Indy loaned by SGI in Newport and color x-terminals loaned by NCD.

The computer network cost, exclusive of salaries was: $1000 Cerfnet, $2223.33 Digital Networks, $1000 meeting room rental for August 5 for Digital to install wiring for a total of $4,223.33

RECOMMENDATIONS:

1. The computer facilities (workstations, network, peripherals) were used to advantage during the workshop. Many people used them for much more than just accessing email. While not the majority of the 140 registrants, there were approximately 30-40 regular users. Approximately 30 people indicated that they would bring laptops to the meeting. 10 to 15 people actually connected their laptops to the network during the meeting. All were very well prepared on how to do this in advance of the meeting.

2. There was no networked PC provided by the workshop organizers at the meeting; in a future workshop this would be very useful.

3. An ftp site for data dropoff well ahead of the meeting was provided. However, most data were brought to the meeting and loaded on the computers there or were accessed over network links.

4. The 128Kbit ISDN line was adequate. Attendees were favorably impressed by its speed.

5. A network for just a central "computer room" would have been enough. All equipment provided in the three central, connecting rooms was used - the workstations, printers, peripherals and hubs to which laptops were connected. The network connections to the remaining conference rooms were not necessary.

6. An onsite network manager is essential for smooth functioning. The manager who supported the Pacific workshop installed the network, connected all machines, answered questions, and made things work smoothly for people during the meeting. Volunteer help was plentiful and is gratefully acknowledged.

7. The SGI Indy Presenter with overhead projector worked well for presentations to an audience.

8. The outside ISDN connection was made well in advance of the workshop, plus wiring to each of the conference rooms. However, the hotel could only allow access to the meeting rooms starting at 8 A.M. on August 13, and also required that everything be removed by 8 A.M. on August 24, which meant that all equipment had to be unloaded and set up and then taken down in a very short time. An extra day for setting up the network would have been useful.

S1. Heat and freshwater fluxes and transports

Estimation of meridional heat and fresh water transport by the ocean circulation is one of the chief goals of WOCE. At the workshop, the problem of determining these fluxes was focused in two topic areas: the exchanges across the air- sea interface, and the advective flux divergences in the ocean. In an internally consistent description of a given ocean volume, the sum of these two equates with the time rate of change of storage. A presentation by P. Niiler exploited this relationship on the annual time scale to help develop and validate new estimates of the surface heat flux in the Pacific. The flux estimates, validated at annual period, yielded new maps of the climatological (time scales where storage change can be neglected) air-sea heat exchange over the Pacific (Moisan and Niiler, 1996). One surprising feature that emerged was a broad area of net ocean warming in the subpolar eastern Pacific, presumably balanced by an advective divergence in the ocean. Widespread interest in this result led to the creation of a working group to further investigate the air-sea heat flux in the NE Pacific, and its implications. (Highlights of that group's discussions appear below.)

Is this new map of air-sea heat flux exchange accurate? It certainly differs from several previous surface analyses in many respects such as the northern region of net warming. Yet the new fields appear consistent with previous ocean flux divergence estimates at Lats. 24 and 10N, 28 and 43S. Also reported in working group were other new flux products now, or soon to be available including stress, latent heat and precipitation. Which of these are considered closer to "reality" than others. More precisely, which are deemed sensible for comparison with, or appropriate for constraining, ocean flux divergence estimates? These questions motivated creation of a working group focusing specifically on comparing and evaluating estimates of surface fluxes. The sub-group discussed the need for up- to-date evaluations of available surface flux fields in order to qualify them for WOCE-related activities. Particularly in light of recent availability of several reanalysis products, most notably from NCEP (formerly NMC) and ECMWF, such an evaluation would have value for selecting appropriate choices (inappropriate choices should also be noted) in constraining flux divergence estimates and forcing numerical ocean models.

Several tools/data sets were discussed as metrics for evaluating the various products: revised climatologies corrected for observation system alterations (e.g. daSilva et al., 1994); independent in-situ data (e.g. WOCE Surface Meteorology DAC); remotely-sensed estimates (Gautier and Frouin, 1992); and derived flux estimates (e.g. Moisan and Niiler, 1996). Ocean-based estimates of meridional heat and freshwater transport, Sverdrup transport, and Ekman transports are also potentially useful for qualifying various flux datasets. Sensitivity analysis employing ocean model(s) forced with different flux fields should also be considered as their results can be compared to independent data such as sea level, measured transports and currents, and heat and freshwater transports from WOCE sections. Some limited intercomparisons and sensitivity analyses are underway. Many of these efforts are focused on the momentum fluxes for the tropical Pacific where in-situ based analyses (FSU) have highlighted the deficiencies and successes of available operational products and the consequential impact on ocean modelling. Remotely-sensed data from ERS-1 and (soon NSCAT) have also proven valuable for evaluation purposes. There are other efforts underway not represented at the Pacific workshop.

A comprehensive evaluation of available flux products will not likely emerge for some time given the diverse and distributed nature of underway activities as well as the resources necessary for such an intercomparison. In the days following the workshop it was learned that two proposals have been submitted (one to WCRP and another to SCOR) to establish a working group to compare various flux products (primarily reanalysis fields), listing their advantages and disadvantages, and issuing recommendations for their use. It is unclear what role WOCE could play in such a working group, but clearly WOCE would benefit from such an effort.

Estimation of air-sea fluxes through ocean advective flux divergence calculations is, in large part, a by-product of determining the ocean circulation. (The basic problem of the Pacific's general circulation was discussed in several working groups during the week.) As was stressed by S. Wijffels in her presentation of heat and water transport estimates at Lat. 10N and 30S, there are several issues key to obtaining accurate estimates of advective fluxes with limited uncertainty bounds. These must not be overlooked when addressing the general circulation question. Large vertical property gradients exist in the upper ocean. The ageostrophic wind driven (Ekman) property transports depend critically on how the ageostrophic motions depend on wind stress, knowledge of the mean wind stress, and in particular, how the velocity is distributed vertically. WOCE data analysis efforts aimed at bettering understanding of the directly wind-driven property transports are encouraged.

Strong flows with anomalous water properties (compared to the ocean interior) are also found at lateral boundaries. Focussed efforts to directly estimate those boundary currents were made at several locations during WOCE, but not all. What data are available in the boundary regions of sections must be carefully examined for their contributions to meridional property transport. By no means of least importance are questions of time variability including seasonality in surface properties and transport. Wijffels noted relatively little difference at Lat. 10N between net property transport estimates based on synoptic and time-mean surface layer properties. Far greater was the uncertainty in property transports stemming from differences in mean wind stress at this latitude between the available climatologies. This may not be true at other latitudes. Also of concern for direct property flux estimation is how representative the synoptic section data are of the time- mean properties. Here the repeat hydrography and high- resolution XBT programs may provide guidance, as well as the altimetric observations of sea level change on the lines. Advance clearly requires widening the scope of analysis efforts beyond an individual line of hydrographic stations.

da Silva, A. M., C. C. Young, and S. Levitus, 1994: Atlas of surface marine data 1994 Volume I: Algorithms and procedures. US Government Printing Office, Washington, DC, NOAA Atlas NESDIS 6, 83 pp.

Gautier, C., and R. Frouin, 1992: Net surface solar irradiance variability in the central equatorial Pacific during 1982 - 1985. J. Climate, 5, 30-55.

Moisan, J. R. and P. P. Niiler, 1996: The seasonal heat budget of the North Pacific: net heat flux and heat storage rates (1950-1990). J. Phys. Oceanogr., submitted.

Jim McWilliams, UCLA

Several new analyses of Pacific surface fluxes and properties have been developed recently. Moisan and Niiler's (1996) new mean annual net heat flux map for the globe is based on their considered best choices of bulk formulae and satellite measurements, as carefully compared with the observed seasonal heat storage cycles in the North Pacific. An apparent weak warming appears in the subpolar regions, especially in the eastern North Pacific; weak warming is also apparent in the Southern Ocean. This somewhat counterintuitive result was discussed in more depth in the heat flux working group and stimulated a separate discussion later in the workshop.

In the surface layer, the non-coincidence of the actually mixed layer, the layer containing ageostrophic flow, and a more generally defined boundary layer may be critical for how parcels are injected into the interior. In the Large model formulation for the upper ocean layer, the North Pacific boundary layer can become as deep as 400-600 meters; the questions are whether this occurs in observations and whether it is possible to use available measurements such as ADCP profiles to look at mixing. The ADCP's vertical resolution (15-20m) is not sufficient in itself to resolve the critical Richardson number, but the question arises of whether it is possible to parameterize mixing at the small scales for the coarser scales which can be measured with the ADCP.

Drifters show that the ageostrophic velocities are Ekman-like, with response depending on the stratification as well as the Coriolis parameter. A drifter product is now available; comparisons could be made with Topex/Poseidon.

Surface energy levels and variability can be estimated using many different techniques as a result of the WOCE and satellite data sets. A useful comparison would be of the ADCP-measured velocity fields and their horizontal wavenumber structure and those measured from TOPEX/Poseidon, drifters, AVHRR SST, and models.

Dan Rudnick, Scripps Institution of Oceanography

The abyssal circulation group met for the first three days of the week. Our first two days were devoted mainly to presentations by several members of the group, and discussion of results. There was sufficient time for 15-20 minute presentations by all who expressed a desire to talk. The presenters were, on Monday, G. Johnson, B. Smethie, J. O'Dwyer, and R. Key, and on Tuesday, D. Rudnick, J. Bullister, K. Mitsuzawa, F. Santiago-Mandjano and D. Musgrave. Beginning on Tuesday, and through Wednesday, we held a general discussion assessing the current knowledge of abyssal flow, and future directions.

We had a very interesting discussion with the goal of identifying what WOCE observations have taught us about abyssal circulation. First, we concluded that the improved temperature, salinity, oxygen, and nutrient data base has provided a much better first order description of the abyssal ocean. Second, the transient tracers provide a valuable new measure of the rates of abyssal circulation. In particular, the CFC signal has moved northward as far as the Samoan passage. We know that this signal has reached this point in only the last few years. Third, the South Pacific current meter arrays at 32S and the Samoan Passage have yielded accurate estimates of the mean abyssal transport, and its variability. The known transports also allow the inference of diapycnal velocities and diffusivities. Fourth, we have new information about the complicated abyssal boundary currents of the North Pacific from current meter arrays and tracers. Taken together, we concluded that WOCE observations have provided an improved description of, and new insight into, abyssal circulation. Approximately eight manuscripts were likely to be ready for submission within a year.

We next discussed important topics for further analysis and synthesis. The WOCE Pacific data base includes a number of measures of abyssal flow: current meters, geostrophic calculations, and transient tracers. An important project would be to rationalize these three measures of rate in the deep western boundary current. Model data comparisons ought to be carried out where the models should be required to reproduce tracer concentrations and transports as measured by current meters. There was widespread desire to use the whole of the hydrographic data set to produce an estimate of the general circulation. A few of the abyssal circulation issues that might be then addressed are: (1) the inference of diapycnal velocity and diffusivity, (2) the abyssal transport of nutrients, and their effect on CO2 uptake and storage, (3) the importance of narrow maxima in tracers and flow along boundaries in balancing large-scale budgets. A number of members of the group began making connections with the target of writing manuscripts on these topics in 2 to 3 years.

Terry Joyce, Woods Hole Oceanographic Institution

The western gyre region is atypical of most of the Pacific WOCE program because of the large number of repeat hydrographic sections being done by Japan, ROC and PRC. These repeat occupations are important for setting the WOCE dataset in context and can directly address questions about the representativeness of the onetime survey (i.e. was WOCE an anomaly?). These studies include observations of Kuroshio transport (hydro lines, current meter moorings: PCM2), surface drifter variability, satellite altimetry, surface fluxes, mode water formation, and intermediate water variability. While some variability can be linked to ENSO signals, mainly in the tropical regions, longer term changes on decadals scale are also being observed.

Modeling studies can be applied to understand whether or not observations made in different time periods can be reconciled to changes in surface wind forcing, for example. Such comparison were made for the western boundary region and flow through Vityaz Strait, but need to be expanded to explain such phenomena as the apparent lack of seasonal signal in Kuroshio transport on PCM2, interannual variability in Kuroshio paths and bifurcations, retroflection of the Mindanao Current in the Celebes Sea and its effect on Pacific-Indian throughflow and eddy formation in the South China Sea near the Luzon Strait. There is a great deal of model/data comparison needed and this workshop established links between those with models and those with data.

Onetime survey work in the region is still ongoing: a recent occupation of WHP line P8 was only just completed before the workshop by three Japanese vessels and additional onetime survey work in the western region by ROC was planned for later this year. Much of the effort of the scientists on these lines is still directed towards writing up initial results from their sections including observations of flow by shipboard or cable-lowered ADCPs. A similar situation exists for the deep circulation observed with floats. Most of the RAFOS results are completed, yet still some Japanese measurements are ongoing as are many of the ALACE floats. Coverage of floats by float type was complementary: ALACES in the south, RAFOS in the northwest region. One obvious need is to combine results of onetime sections with float measurements and current meter results in order to obtain absolute transports: this work is already underway in the NW region as most of the RAFOS results are in hand and the hydrographic program is now complete.

Synthesis of results is already underway as WOCE observations are being combined/compared with pre-WOCE data for water mass studies (NPIW, mode waters), although better sharing of WOCE data would facilitate this process. Many of the PIs involved in this type of effort are beginning to make these exchanges and hopefully the workshop will further aid in this effort. Participants indicated about a dozen or so papers that were underway and would be completed within a year's time. A similar number of longer term projects were identified that were more of a synthetic nature.

Susan Hautala, U. Washington

Session 2, August 20. Joint session with overall and western circulation groups. Presentations from the following eastern and central group participants were made: R. Lukas, A. Flosadottir and J. McClean.

Session 3, August 21. Focus topic: The broad eastern boundary region of the North Pacific. Presentations were made by T. Strub, T. Chereskin, W. White, S. Meyers, L. Zamudio, and R. Nunez.

The connection between the coastal regime of the eastern boundary and the broader scale circulation in the eastern gyres is a relatively underexplored problem. Many recent studies have begun to examine westward propagation of variability away from the eastern boundary at a variety of frequencies and spatial scales. One particularly interesting problem was identified at the annual frequency. Sea surface temperature and sea surface height show a seasonal generation of eddy energy in response to changing wind regimes a long the eastern boundary, which begins to propagate westwards but apparently ceases several hundred kilometers offshore. Observations made by the participants of this working group were discussed and Dudley Chelton also addressed the phenomenon in a separate keynote lecture (August 22). A loose association of investigators (Chereskin, Donohue, Kelly, Meyers, Nunez, Paduan, Strub, White and Zamudio) formed to study this phenomenon and the more general question of how the eastern boundary region affects properties of flow in the interior. An important part of this study will be the examination of subsurface temperature data - observations thus far have been restricted to the surface expression of the eddies. Related questions include identification of the true critical latitude for annual Rossby wave propagation, characterizing and explaining the width of the energetic boundary current region, explaining the apparent dissipation of eddy energy within 1000 km of the eastern boundary and determining the impact of the eddy life cycle on mixing of properties and momentum. Interested parties should contact T. Strub or T. Chereskin about plans for future meetings.

Related questions of interest to the group include the eastern boundary response to subannual forcing (i.e., is the observed eastern boundary connection to high latitudes made via Kelvin wave propagation or response to local forcing in the atmosphere?) and a comparison to the South Pacific that will start with an analysis of the basic structure and variability of the southern hemisphere eastern boundary system.

Session 4, August 22. Focus topic: the central and eastern interior subtropical gyre. Presentations were made by L. Thompson (w/ S. Hautala), and T. Suga.

WOCE and other datasets are revealing that the picture of subduction and ventilation in the subtropical North Pacific is more complex than expected. Besides the classical mode water of the western Pacific, distinct mode waters have been identified in the central and eastern subtropical gyre as well.

Analyses of archived XBT data, which is particularly useful in the Pacific because of the relatively shallow thermocline, are revealing distinct decadal changes in subsurface thermal structure. In a separate keynote lecture on Aug. 21, Kimio Hanawa proposed an intrinsic mode of variability connecting the known atmospheric response to low latitude thermal anomalies to the eventual return of higher latitude thermal anomalies induced by atmospheric teleconnection patterns to the tropics via subduction and gyre-scale advection in the main thermocline. We identified a group of investigators working on decadal changes and on the basic structure and seasonal variability of the Pacific ventilated thermocline (Hanawa, Hautala, Kelly, Qiu, Suga, Thompson, Watanabe, White).

A small group interested in low-latitude circulation was formed during the workshop. During the first two of its three meetings, the group discussed recent and ongoing analyses of low-latitude data sets. The last meeting focused on plans to to bring together disparate data sets to address two issues of particular interest: the upper and intermediate circulation in the low-latitude western Pacific, and the system of predominantly zonal subthermocline currents in the tropics.

During the first meeting, Tomowo Watanabe introduced the data from a Pacific-wide cruise along the equator on the Kaiyo Maru in January to March, 1992. This unique data set includes upper ocean velocity measurements from a 75-kHz ADCP, and full-depth CTD profiles. Yuji Kashino showed data and analyses from several cruises in the far western equatorial Pacific, from 1992 through 1996. Julie McClean discussed her work with the Los Alamos parallel ocean program (POP) model, and in particular her interest in comparing the model with observations in the equatorial region.

The second meeting continued the western Pacific theme with Fred Bingham's presentation of his analysis of hydrographic variability in the long time series of cruises on 137 E. A second theme was also introduced, as Greg Johnson and Lynne Talley spoke about the evidence for deep zonal currents from T/S and oxygen anomalies seen in the WHP sections and in isopycnal property maps including the WHP data. Lynne Talley also showed results from her ongoing analysis of Brunt-Vaisala frequency profiles from the WHP sections, pointing out the presence of a high-stability layer overlying a well-mixed layer at the bottom.

Observed and studied are: diapycnal/isopycnal mixing rates and vertical velocities; biological processing rates (O.U.R.s), nutrient fluxes and pathways; water mass formation rates; dilution rates and scales; intergyre, cross equatorial, cross ACC exchange.
The ventilation and mixing working group set a schedule covering

1. Mode water formation and thermocline ventilation on Tuesday: subduction and other processes; comparison of tracer ages with formation rates; intergyre exchange & feeding; tropical upwelling; oxygen minima and nutrient traps.

2. Intermediate water ventilation and formation on Wednesday: The NPIW question: can we answer it?; Subtropical <==> subpolar exchange; AAIW formation and circulation.

3. Deep and abyssal ventilation and formation on Thursday: Deep water signatures; AABW intrusion into the South Pacific.

4. Generalities on Friday: concepts of tracer ages vs. ventilation rates vs. circulation; assimilation of tracers into models; and interpretation.

General Recommendations:

1. That models be run with appropriate tracers suitable for evaluation of both model performance, and as a guide for interpretation of the tracer fields.

2. There exist observational diagnostics of model performance (regarding ventilation and intergyre exchange). These include:

3. Improvement in boundary conditions may be necessary: this includes both improvement in understanding of how tracers actually are introduced to or behave within the ocean, and also how the models handle these aspects.

4. We need to move beyond the initial descriptive, and semiquantitative phase of data presentation.

5. Synthesis of different tracer techniques needs to be accomplished.

6. Collaborative efforts should be made to combine and improve interpretive skills between hydrographers, tracer geochemists, modellers, floaters, etc.

Specific action items for long term projects:

NS1. Subpolar gyre - circulation and heat balance

NS2. Seasonality of the Kuroshio transport

A Working Group newly established during the workshop met on Thursday morning (Aug. 22) to discuss seasonal signals of the Kuroshio transport. About twenty people attended the meeting and had very active and informative discussions. Shiro Imawaki (Kyushu Univ., Japan) presided the meeting. He explained briefly why this WG was established; the transport of the Kuroshio south of Japan, estimated recently, does not show an apparent seasonal signal, which fact is somewhat different from previous works including altimetry data analyses and numerical model studies.

At the beginning, nomenclature for the western North Pacific was briefly explained. Then, the present status of our knowledge about the Kuroshio transport was reviewed by a series of short presentations, including talks by Shiro Imawaki, Kimio Hanawa, Bo Qiu, Dudley Chelton, Jim McWilliams, Harley Hurlburt and Paul Myers. Chen-Tung Chen and Rongfeng Li also presented short notes on circulations in the South China Sea and Philippine Basin. Summary of discussions is as follows. The seasonal signal of the Kuroshio transport is or seems to be small in the East China Sea and south of Japan, where the Kuroshio flows along the western boundary. For the Kuroshio Extension, after leaving the boundary, the seasonal signal of its surface transport is somewhat apparent, according to an altimetry data analysis, but the seasonal signal disappears when the transport associated with the recirculation on the southern side is removed.

A session was held devoted to the interpretation of the TOPEX/POSEIDON data in terms of its implications for the deep flows. From comparisons with current meter moorings, a slightly over-simplified conclusion is that surface kinetic energy calculated from the altimeters reflects primarily the first baroclinic mode of motion. Thus one has the elegant result that the altimeters are displaying primarily the motion of the main thermocline, and we have a window into the deep interior of the ocean. There are, however, regions, e.g. close to the Gulf Stream, where the motions are dominantly barotropic, and various evidence suggests that high latitudes are increasingly barotropic in nature. But there are few high latitude current meter moorings suitable for this purpose. Near the ocean edges, higher modes do emerge as important.

Comparisons with general circulation models support the inference of high latitude barotropic dominance, particularly in the elevation itself rather than in the kinetic energy. Analyses have not advanced sufficiently to produce clear understanding of the model reliability for these purposes. That is, real differences between altimetry and models have emerged and the vertical partitioning within the models may not yet be accurate. Work on model/altimetric data comparisons is continuing vigorously in a number of groups. The best results are expected to emerge when models are actually constrained to the altimetry.

More work needs to be done in comparisons of the altimetry with inferences from the repeated XBT lines. There was call to greatly expedite availability of the complete XBT data sets. Pacific ATOC data should produce a novel test of the understanding of the altimeter derived motions.

Motivated in part by the relationship between the ocean's general circulation and the meridional heat and water transports effected by that circulation, a subset of meeting attendees associated with the WOCE Hydrographic Program one-time survey and the high-resolution XBT program met to discuss progress with analysis of the individual lines and plans for synthesis. Short presentations were solicited from representatives of each of the WHP lines in the Pacific, who were also asked to estimate their time to complete and write up an initial description of their observations. A brief summary of these presentations follow. Several topics of common interest to researchers were found, including the deep silica distribution and the stratification and circulation near the equator, that led to creation of other working groups.

P1W: H. Freeland - two papers describing these observations have been revised in light of journal referees comments and have been resubmitted.

P1E: D. Musgrave and H. Freeland are using these data in combination with repeat and other 1-time hydrographic sections to investigate the circulation in the NE Pacific

P2: I. Yasuda discussed mode water properties along this line and related the observations to the talk given in plenary by K. Hanawa. The group of P2 investigators intends to have a descriptive paper on the section completed in approximately 1 year.

P3: pre-WOCE section completed in 1985

P4: pre-WOCE section completed in 1989

P5: not subscribed

P6: S. Wijffels reported on heat and water transports on this line in plenary. M. Tsimplis reviewed the inverse modeling work of the SOC group using this line and P21. Both groups anticipate initial papers will be completed within one year.

P7: historical section (Scorpio) Australian's have reoccupied that Tasman Sea segment and have published thermocline-depth water mass shifts.

P8: H. Ishii reviewed the multi-ship program directed by N. Suginohara that recently sampled this line in June-July 1996. The data are presently being calibrated.

P9: I. Kaneko discussed the silica distribution along the P9 line, and showed a section of zonal velocity to ~800 m depth observed with a low-frequency ADCP. This group plans to use the direct velocity data in combination with the hydrography on this line and P24 to develop a description of the circulation in the Philippine Sea. M. Ishii showed preliminary observations of del-C14 on the P9 section. Papers will be completed in 1- 2 years time.

P10: T. Joyce and S. Wijffels showed data collected on the P10 line and summarized plans for publication. Their initial focus has been on the region north of Lat 10N bounded by the P10 line and the pre-WOCE P4 data. A manuscript is anticipated by year's end. On a roughly 6-month time scale, joint work by Joyce, E. Firing and P. Hacker on direct velocity observations in the Kuroshio will be completed. Subsequent research will focus on the equatorial parts of P10.

P11S: no representative of this line was present at the meeting

P11N: not subscribed

P12: no representative of this line was present at the meeting

P13: M. Kawabe highlighted potential vorticity and dissolved oxygen features in the thermocline along P13. This group is currently writing a paper on the water mass properties and expect to complete it in one year. Another paper on the deep water is anticipated. G. Johnson called attention to a thin filament in the silica distribution at the northern boundary of their occupation of this section. He reports that B. Taft is focusing his attention on the subpolar parts of the section.

P14N: G. Roden called attention to the role of bathymetry in controlling the circulation on the P14 line. A description of the Bering Sea crossing was published last year. Future papers will include a description of the upper layers within 1 year, properties at depth in 1-2 years, and the direct velocity observations within about 6 months.

P14S: G. Johnson's initial interest in this and the P15S data is how the deep waters negotiate the turn around the Campbell Plateau, moving from part of the surface intensified ACC to a bottom-intensified deep boundary current.

P15N/S: H. Freeland anticipated a joint paper discussing the properties and circulation on this very long combined section in about a year. He also highlighted the detailed sampling conducted about the exit of the Samoan Passage that was suggestive of hydraulic flow.

P16: L.Talley summarized her regional interests that utilize the data from this and the associated legs in the SE Pacific. Topics include AAIW properties and circulation, equatorial jets, abyssal flows and the role of near-bottom processes, and communication between the tropics and subtropics.

P17S: L. Talley - initial paper published on the section. see above

P17N: D. Musgrave is working with J. Lupton on the deep circulation in the NE Pacific with guidance from the deep silica and 3He distributions. He is also interested in the oxygen minimum waters, and has experimented with local box models using inverse techniques. He expects to have an initial paper describing the section and water properties drafted by the end of 1996.

P18: G. Johnson has used these data in his joint analysis with Lynne Talley on the deep zonal flows in the Pacific. Given the close distance of this line to P19, he does not believe a water property descriptive paper for the line is needed. Greg called attention to observations of small vortices at the southern end of the section observed with the LADCP and seen in the density stratification.

P19: M. Tsuchiya and L. Talley have a preprint describing this section in hand.

P21: M. Tsimplis and colleagues at SOC are investigating the subtropical S. Pacific circulation with a box model and inverse techniques applied to this section and P6. A paper will be submitted within 1-2 months. M. Baringer and M. McCartney have been using these data in combination with other WHP observations to document the properties and circulation of the AAIW.

P24: I. Kaneko and colleagues are using this line in combination with P9 and P3 data to investigate the circulation in the Philippine Sea.

P31: D. Roemmich and colleagues have a paper describing these data in press.

S4: G. Panteleev presented a synthesis of WOCE era and historical hydrographic data in the far South Pacific, carried out by he and colleagues at the Shirshov and Moscow Institutes. The analysis used variational methods to assimilate the hydrographic data and return an estimate of the absolute circulation.

M. Morris reported on her analysis of repeated, high- resolution XBT observations that enclosed the warm pool waters of the western equatorial Pacific. Her study of the heat and water budgets of this region will be written up very soon. J. Sprintall showed results of a study of the Tasman Sea circulation and heat budget that also used high- resolution XBT's. This work, part of which has already been published, highlights the importance of advection in the heat budget of this region. Lastly, L. Talley gave a verbal description of upper ocean variability along 30S observed by high-resolution XBT's, and possible implications for interpreting the P6 one-time section data.

Subsequent discussions by this group were aimed at developing a synthesis of the WHP and HR-XBT observations. It is expected that several groups will be involved with such efforts in the coming years. One idea discussed was to bring together the investigators who have been intimately involved with collecting and interpreting data on individual lines to develop a basin-wide estimate of the 3-D circulation. This will take the form of multiple-box models of the Pacific using the one-time survey data to subdivide the ocean, and historical observations (chiefly the Reid edited data set) to document properties within boxes. G. Johnson, S. Wijffels and J. Toole will lead this effort initially, but they are inviting (and expecting) significant input from the WHP investigators. Formal plans to propose this effort were discussed.

T1. Data-based climatology

This group met as a whole on the first day of the meeting. It was clearly too large a group for general discussion and so it was decided to split into two groups on the second day. The reports of the two subgroups follow.

The groups were composed almost exclusively of observationalists. Data requirements from the modeling, model/data comparison and assimilation groups were not incorporated in the following reports.

1. Hydrographic data
1.a. Data availability and access

A large amount of Pacific WOCE data was assembled by the WHPO prior to the workshop, including large amounts of tracer data, including most of the CFC data. This momentum should be maintained, to assemble all the available data.

Data are available at the workshop through the WHPO electronic atlas. These will also be made available after the workshop to those investigators who have sent in their data. Which data should be made available to whom after the workshop?

i) All publicly available data will be available for everyone.
ii) Semi-proprietary data can be password protected.
iii) The files in the atlas do not need to be the complete files, which are maintained separately elsewhere. Specific data columns thus do not need to be released along with other data.

It was suggested that a more usable format be used for distributing data if desired, for instance flat ascii tables including both station and data information, for instance, or the same in netCDF files.

It is recommended that the WHPO assemble the pre-WOCE tracer data, including for instance Lupton's helium data, and Weiss and Warner's CFC data from P1 and P3.

It is recommended that some historical data be converted to WHP format. The principal candidate for this is the Reid/Mantyla selected data set.

It is recommended that NODC use uniform data formats, retaining all station information and precision.

1.b. Hydrographic data products

1.b.1. Crossover differences for hydrographic data: statistical check on differences between sections (Aoyama, Mordy/Gordon).

1.b.2. Uniform data set with adjusted S, O, nutrients based on crossover differences (Aoyama; Joyce; Mantyla)

1.b.3. Uniform tracer data set with age-adjustment for tracers (up to 2-2.5 years, a 10-20% change). For model validation suggest that models compare with the time when the data were actually collected. (CFC - Warner/ Bullister; Tr/He3 - Jenkins, Top; C14 - Aoyama, Ishii, Key)

1.b.4. Gridded sections: currently being produced by Orsi, Talley and Freeland independently. Suggest discussion amongst them on methods (comparison) and discussion with SAC about inclusion of complete gridded section set in the electronic atlas. Also could make available using EPIC.

1.b.5. Gridded 3-D hydrographic data (T, S, O2, nuts):

The WHP SAC currently has climatology now for 45S and farther south, and is working on climatology to 20S. Working to equator. Data are quality controlled. A quality-controlled data set is available. Data are adjusted and then the climatology produced, using isopycnal averaging, with an e-folding scale of 500 km. The data separation is O(150 km) with the full data set, much sparser if only quality-controlled data are used. (Gouretski)

Hydrobase: quality control on N. Pacific data (Macdonald and Suga) To be completed by early 1997.

Recommendation: Gouretski, Macdonald and Suga meet together and compare quality control methods. SAC qc the S. Pacific and Macdonald/Suga the N. Pacific. Use qced data sets for both databases. Later on compare the two methods. SAC might make their database more interactive. "Mean" climatology (WOCE best choice of parameters and gridding) and interactive databases available on CD-ROM.

Suggestion: for the upper layer, seasonal mean data should be computed. It is not clear who will do this.

1.b.6. Gridded 3-D tracers and maps - use very coarse final spacing and a single point in time for age adjustment for the transient tracers. Map pre-bomb C14 and He deeper than about 1500m. (Smethie, Bullister, Warner, Orsi)

1.b.7. OceanAtlas for PowerMac is available with available Pacific WOCE data. (Swift)

1.b.8. Movies: movies are already available for the 137E data (Bingham), maps of temperature (Diggs, White at JEDA center), some XBT sections (Morris, Sprintall)

2. Low resolution XBT data

2.a. Data access and availability

Delayed mode data through 1993 are at NODC, and also are available through the JEDA Center's GoodBase (a branch of NODC).

Near real-time data through July, 1996, not quality controlled, are also available at the JEDA Center, updated monthly.

NODC in Silver Springs maintains the archive and can provide quality- controlled data.

Users currently can receive the data in several formats, including ascii. netCDF takes a lot of disk space, and so is not a first choice for a uniform data format, but it could be contemplated.

2.b. Data products

Currently the JEDA Center produces gridded fields based on the broad-scale XBT's. These include a 10-year climatology (1980-1989) at standard levels (0-400m), and monthly maps. It covers the region 30-60N on a 2 x 5 grid.

3. High resolution XBT data

3.a. Data availability and access

Publicly available data are at NODC and include < 1 month GTS, 1 degree subsampling and the delayed mode data up through 1993.

All data along 7 transects through 1996 are available through the principal investigators at SIO (Roemmich, Sprintall). Vertical sections of temperature are available from them through ftp.

XCTD data are not yet quality controlled and so have not been released.

3.b. Data Products

All temperature sections are gridded within 2 months of collection. These are available on a website. (Roemmich, Sprintall)

Bathymetry: attempt to make the latest high resolution version available on the CD-ROM?

4. Distribution of data and data products

Printed media: there appears to be a split opinion on production of printed atlases. A poll of the plenary session at the conclusion of the workshop indicates that there is overwhelming support among the scientists for printed atlases. Full color, hard cover volumes would probably be prohibitively expensive. The suggestion is that reasonably cheap atlases (similar to soft cover data reports) be produced for the Pacific 1-time survey.

A printed atlas (data report) for the high resolution XBT sections is currently being produced.

CD-ROMs: This will be the principal means of distribution on a permanent medium, but there is a need for migration. The WOCE DIU will release a CD-ROM with all currently available data next spring. A later release could include the WOCE "climatologies".

It is possible that NODC would release CD-ROMs for no cost.

Webpages: interesting and increasingly useful, but transient. These are the medium of choice for evolving products and for easy, current, interactive access to data and products. However, it is unclear whether these should be considered the permanent distribution.

1. Data availability. There is concern about the much delayed availability of the final XBT results. NMC does quality control of a subset of the real time data, and makes it available by ftp. Ming Ji is a contact. But the latest products from NODC end in 1992 and this is not acceptable. WOCE needs to do something. (See recommendations and expected timeline in the recommendations section I.1.)

Surface meteorology fields are available via NCAR. There is a six week delay on NMC products and longer delays (subject to financial difficulties) between ECMWF and NCAR). NRL archives the FNOC fields which are believed by some to be better than anything else operational in the tropics; Harley Hurlburt is a contact. As to the known reanalysis products (NMC, ECMWF, DAO-Goddard), no one is prepared to recommend one as best, and it will probably be some years before any consensus will emerge. So no recommendation is made on a surface meteorology reanalysis product.

Otherwise, data availability did not appear to be a first- order concern, or at least not much different from the wider worries about proprietary rights. Multi-author and credit problems are part of the wider WOCE concern.

2. Special programs required, etc. There was a strong consensus that highest priority is now to produce PRODUCTS of assimilation. While there is much to be learned yet from new mathematical and algorithmic approaches, these should take a definite second priority to actually producing syntheses of real data with GCMs. We know enough to do a great deal in a practical way with methods which are well understood - fundamentally least-squares - and these should be exploited.

Related to this is a the consensus that we have too few people who are knowledgeable about both models and the realities of data. Programs such as the OSU summer school will help, but unless people return to groups which actually carry out estimation, much of the benefit will be lost.

The practical translation of both these paragraphs is that the funding agencies need to focus more on practical WOCE synthesis and somewhat less on the interesting mathematical theory of estimation and control. We have adequate review papers and text books in this area; we now need experience.

3. Resources. Enough is known of assimilation to recognize that desirable methods and data will swamp the largest conceivable computers for many years to come. We need to be clever about what we do with the available power. There are serious concerns about whether the assimilation community will have adequate access to even the present generation of computers. In particular, supercomputers available to the oceanographic community tend to be balkanized into many small jobs, rendering it difficult, if not impossible, to run large jobs. Some form of "fencing" is necessary.

4. Special issues and papers. Progress in assimilating the WOCE Pacific data sets is not adequate to produce true combined products in the next year. Little or nothing can be done about this, as much of the data are just now becoming available, and the community cannot move much faster than it is already. Within 2 - 3 years however, there should be substantial results. The preliminary model/data comparisons are and will be available on the shorter time scale, and should be regarded as part of the assimilation effort.

5. Model/data comparisons. Specific model/data comparisons for assimilation are thought best left in the hands of those attempting assimilation as the best judges of what are the critical points. So no general recommendations are made on this topic.

6. We had no information on specific requirements on model products required by the "observational" community.

The concurrence of progress in several areas puts the WOCE community in a favorable position to achieve a qualitative improvement in the modeling and dynamical interpretation of the large-scale circulation of the Pacific and other ocean basins. This progress includes, first, the successful acquisition and present availability of the WOCE datasets, both in the Pacific and elsewhere. A second such factor has been the dramatic evolution and improvement in the quality and variety of oceanic general circulation models available to the international WOCE community.

Together, these new datasets and models offer the exciting prospect, largely unachieved previously, of conducting quantitative model/model and model/data comparisons for the large-scale ocean circulation. The goals of such systematic comparisons include the validation and improvement of models; the quantitative assessment of alternate numerical methods, physical parameterizations and/or simulation procedures; and the dynamical synthesis and interpretation of the WOCE observations.

Many examples were presented of sensitivities in model performance due to alternate choices of numerical formulation (e.g., vertical coordinate, spatial resolution, and others), subgridscale parameterization, and/or environmental fields (e.g., bathymetry). It is now becoming increasingly clear that, far from being of secondary importance, such considerations as these have zero-order consequences for model fidelity.

An important next step for ocean circulation modelers will be the identification, adoption, and implementation of a consensus set of idealized two- and three-dimensional model test problems. Such problems should be defined relative to oceanic processes or properties which are independently understood (perhaps analytically), and should emphasize attributes of known importance to large-scale dynamics (e.g., effects of rotation, meridional boundaries, steep topography, strong stratification, etc.). There should be objective and unbiased standards of success. Lastly, the test problems should not be prejudiced for or against any particular model algorithm or approach.

A start has been made on the formulation of such a suite of test problems. With support from the Department of Energy (DOE) High Performance Computing Center (HPCC) Program, a small workshop was held in Santa Fe, New Mexico in September 1994 (Test Problem Working Group, 1996). Attended by 11 scientists from the U.S. and international academic institutions and DOE national laboratories, the workshop identified a hierarchy of seven test problems. These tests address such process-oriented issues as the propagation of an equatorial Rossby soliton, advective conservation of both dynamic and passive scalars, steady nonlinear geostrophic flow about a tall seamount, steady western boundary current structure, adiabatic wind-driven flow in a mid-latitude basin, and form stress effects in a simulated Antarctic Circumpolar Current. Three additional test problems, dealing with flow near continental boundaries, have been proposed and explored numerically by Haidvogel and Beckmann (1996).

1. Eddy Resolving Models

Significant theoretical effort in WOCE has been in developing and running basin-scale, eddy-resolving models of the general circulation of the oceans. Many of these model integrations have been with time dependent forcing, from seasonal to synoptic time resolutions. Pacific WOCE data sets contain significant information on the time variability and intercomparisons should be made of the time variability of this data with the comparable eddy resolving model parameters. The objectives are to assess both the energy level of the variability in the models as well as the role these variabilities play in the general circulation of the oceans.

In the Pacific, the variability can be divided into a mesoscale (motions with periods of 10-200 days), seasonal cycle (periods of 6 months and 12 months) and "secular" or interannual time scales (with periods greater than 400 days). In the model-data intercomparison studies, an effort should be made to make comparisons in these separate period bands. Similarly, each of these bands of variability have different vertical structures and horizontal distributions. An effort should be made to quantify these structures and distributions from the modes and from the data.

Most of the model-data intercomparison studies will be done by individuals or small groups of investigators. However, in some instances, modelers will wish to have the data in a readily accessible format and the data analysis will with to have the model data for their use. A variety of diagnostic tests can be constructed, and while there was no agreement on all of these, several commonly agreed upon parameters emerged as being consistent with the original objectives of WOCE modeling and observations. Some data sets are in common use format, while others can be unsed only on cooperative science basis.

2. Common Data Sets and Model Output

i) Sea Level Variability. A large number of investigators are working on comparisons of their model output with the TOPEX/Poseidon and coastal and island seal level data. Results of these projects are being published and a number are planned.

ii) Variability of circulation and mass fields. The computation of the variances of velocity components, temperature, salinity and their covariances should be made from all the data records and from the numerical models. Available potential energy variances should be computed from XBT data.

iii) Heat, Salt and Mass convergences. Where horizontal fields of variability are available their gradient fields should be estimated. This allows for the computations of advection of thermohaline properties, momentum, etc. and horizontal convergence of mass.

3. Commitments

Commitment from modelers to produce fields of variability and their statistics is an individual effort since there are no WOCE modeling centers. WOCE Data Assembly Centers should be asked to provide the following:

i) Sea Level: JPL and Univ. of Hawaii Seal Level Center adequately provide data to users in any requested format.

ii) Drifter Data Center/ AOML and/or SIO will provide all the required parameters from the Pacific Ocean drifter data sets.

iii) JEDA should be asked to provide the statistical fields for the subsurface temperature data. High resolution XBT data is available from SIO on a science cooperative basis.

iv) Float data is available from the WHOI Float Data Center. Statistical fields are available on a science cooperative basis.

v) Current meter data is in the OSU Current Meter Data Center. No funding has been given for massaging this data into required format.

Historical files of XBT, current meter data and drifter data can be very useful in analysis of the WOCE models. No funding or plans for obtaining such funding has emerged for organizing these data sets and for providing these to the modeling community.

4. Simulation and Analysis of the Seasonal Cycle

The suggestion that an Ocean Model Intercomparison Project (OMIP) be organized prompted lively discussion among the working group members. Such a comparison has recently been carried out for atmospheric models (the so-called AMIP; Gates et al., ????). Though the focus of such an OMIP is presently undetermined, its goals would presumably overlap those stated above - i.e., quantitative assessment of model performance, and the identification of factors which improve (or not) model accuracy.

1. Abyssal circulation
2. Western gyre regions and Pacific marginal seas
3. Eastern boundary currents; Gulf of Alaska; eastern subtropics
4. Tropics
5. South Pacific subtropical and ACC
6. WHP (section descriptions or tracer data descriptions)
7. General circulation
8. Heat fluxes (subsume in a surface layer group? WHP descriptions?)
9. Ventilation and mixing, including mode and intermediate waters
10. Model/model and model/data comparisons; assimilation
11. Data and data products (generally reports, atlases, online products)
12. Published or submitted papers (listings from workshop only)

Journal requirements were discussed at length by the workshop committee. The following requirements are considered essential:

i) Established, good track record in publishing special issues in a timely fashion.

ii) Electronic submission of text possible to reduce galley proof errors. Typesetting by the journal available for those who cannot submit electronically.

iii) Color capability.

iv) Capability to collate all WOCE papers submitted to that journal, whether published in the special groupings or not, in a special annual or biannual WOCE volume.

v) Wide circulation of the journal at a reasonable subscription cost.

Appendix A. Agenda

SUNDAY AUGUST 18, 1996

1600-1800 Registration I
1800-1930 Meet and Greet Cash Bar The Garden

MONDAY AUGUST 19, 1996

0730-0830 Registration
0830 Welcome and Introduction - Lynne Talley
0900 Invited Speaker - Carl Wunsch "From the WOCE Field Program to the Synthesis"
1000 Coffee Break
1015 Science Working Groups
1200 Lunch Break
1300 Data Summaries
1430 Technical Working Groups
1600 Coffee Break
1615 Invited Speaker - Bruce Warren "Recent direct velocity measurements in the Pacific deep western-boundary currents"
1700 Model Summaries
1800 Adjourn

TUESDAY AUGUST 20, 1996

0830 Announcements
0845 Invited Speaker - Russ Davis "Absolute circulation in the Pacific"
0930 Invited Speaker - Peter Niiler "New maps of surface heat flux and ocean transport in the Pacific"
1015 Coffee Break
1030 Science Working Groups
1230 Lunch Break
1400 Technical Working Groups
1600 Coffee Break
1615 Invited Speaker - Susan Wijffels "Heat and freshwater transport in the Pacific Ocean"
1800 Adjourn
1930-2200 Technical Working Group - T1. Climatology

WEDNESDAY AUGUST 21, 1996

0830 Announcements
0845 Invited Speaker - Kimio Hanawa "Interdecadal variability in the North Pacific Ocean: subduction oscillation"
0930 Invited Speaker - Jim McWilliams "Modeling and data analysis of large-scale, low-frequency variability in the Pacific Ocean"
1015 Coffee Break
1030 Science Working Groups
1230 Lunch Break
1400 Technical and Science Working Groups
1600 Coffee Break
1715 Invited Speaker - Bill Jenkins "Transient tracers in the Pacific"
1800 Adjourn

THURSDAY AUGUST 22, 1996

0830 Announcements
0845 Invited Speaker - Joe Reid "Circulation of the Pacific Ocean"
0930 Invited Speaker - Harley Hurlburt "Eddy-resolving models of the Pacific Ocean north of 20S with application to the Kuroshio/Oyashio current system"
1015 Coffee Break
1030 Science Working Groups
1230 Lunch Break
1400 Technical and Science Working Groups
1600 Coffee Break
1715 Invited Speakers - Dudley Chelton/Lee-Lueng Fu "Altimetric studies of ocean circulation"
1800 Adjourn
1900 Conference Dinner

FRIDAY AUGUST 23, 1996

B.1. General guidelines

1. Each day's meeting will have a specific focus, to be decided in advance by the working group leader with input from the group. We will be making daily decisions about how to proceed on the following days.

2. There will be time for short presentations relevant to the particular focus by working group members who wish to do so, and ample time for discussion. The working group leader will arrange a schedule for these presentations in advance of the meeting and inform the members.

3. These sessions are primarily intended for discussion. The size of the working groups is a little too large for completely free discussion, so at the discretion of the working group, smaller groups might break off to discuss particular items.

4. Some working groups might meet together depending on the topic.

Shiro Imawaki, Kyushu University, Research Institute for Applied Mechanics, Kasuga 816, Japan

William Jenkins, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA

Steve Rintoul, Division of Oceanography, CSIRO, Hobart, TA 7001, Australia

Steve Riser, U. Washington, School of Oceanography, Seattle, WA 98195, USA

Detlef Stammer, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139-4307, USA

Nobuo Suginohara, Center for Climate System Research, U. of Tokyo, Tokyo 153, Japan

Andrew Weaver, School Earth and Ocean Sciences, U. Victoria POB 1700, Victoria, BC V8W 2Y2, Canada