The U.S. GLOBEC Georges Bank Program Data Workshop

5-12 November 1996
New England Conference Center
University of New Hampshire
Durham, NH 03874

Edited by Peter H. Wiebe
Session reports written by designated Rapporteurs


The support for the preparation of this report was provided by the National Science Foundation and the National Oceanographic and Atmospheric Administration. A special thanks to Bob Groman and Chip Clancy of the U.S. GLOBEC Program Service Office for a superb job in providing logistical support for the meeting and to the New England Center staff for providing a wonderful site in which to hold it.

Table of Contents

I. Introduction

II. Narrative

III. Working Group Reports

IV. Cruise Planning Meeting V. Data Management Discussion VI. Planning for Phase III

VII. References

APPENDIX I. Meeting Agenda
APPENDIX II. List of Workshop Participants.
APPENDIX III. U.S. GLOBEC Phase I Physical Oceanographic Measurements.

I. Introduction

The U.S. GLOBEC Georges Bank Scientific Investigators' Workshop was held from 5 to 12 November 1996 at the New England Center of the University of New Hampshire. This was an intensive, eight day, data analysis and synthesis workshop. The format of the workshop was informal and involved use of networked computers supplied by individual investigators. The goals of the workshop were to:
  1. Conduct analysis and synthesis of data sets.
  2. Enhance Science Investigator working relationships.
  3. Develop ideas for joint publications.
  4. Plan for 1997 field, modeling, and retrospective data acquisition and analysis.
  5. Work to improve the on-line data and information system.

In the words of Cabell Davis:

"... the purpose of the meeting is not to present synthesized results, but to actually do the synthesis at the meeting, while working collaboratively."

In addition, a new executive committee was elected into office during the course of the workshop.

II. Narrative

The workshop began during late afternoon of 5 November 1996 after most of the participants had completed setting up their computers on the local network. Opening remarks by Peter Wiebe described the objectives and goals of the workshop, reviewed the agenda and structure of the workshop, and presented some ideas about possible products resulting from activities at the workshop (a workshop report, collections of papers to be submitted for one or more special volumes, etc). Participants were encouraged to think about the targeting of upcoming scientific meetings that could be used as forums to present the results of the workshop and new findings resulting from this research program. Three meetings described as appropriate forums were:

The latter two were cited as providing the best forums primarily as a result of their timing (the TOS meeting comes in the middle of the very intense 1997 field season), and the nature of the scientific communities that would be attending the meetings. The ICES Annual Scientific Meeting provides a special opportunity for GLOBEC scientists working on Georges Bank to present and compare their research findings with GLOBEC scientists working in the eastern North Atlantic Ocean. A special theme session entitled "GLOBEC: Results from Inter-Disciplinary Programs in the North Atlantic" has been officially approved by ICES. Additional details about the general call for papers will be issued about mid-January 1997. The Ocean Science Meeting has traditionally been an excellent forum for group presentations to the U.S. oceanographic community and it comes at a time when much of the work from the first years of the Georges Bank field and modeling work should be ready for national scrutiny.

Discussion then focused upon defining the disciplinary and interdisciplinary working groups in order to get the workshop running. The following disciplinary groups were suggested:

The following interdisciplinary topics were suggested for group discussions.

On Day 2 (Wednesday), the meeting started with a plenary session (chaired by P. Wiebe) to review the disciplinary session topics and to select the chairman and Rapporteurs for each group. There were three principal working groups formed along disciplinary lines, a physical oceanography (PO) working group which addressed a number of different topics, a predation working group, and a combined vital rates/population distribution and dynamics working group. Bob Beardsley gave a short description of the PO group session organization prior to the groups breaking out in the working group sessions. The plan presented was to discuss physical data collected on Wednesday, modeling the physical data on Thursday, and modeling of the coupled physics and biology on Friday. For the first full day's discussions, the following groups met:


Phase I PO Results Group

Bob Beardsley (Chair), David Mountain (Rapporteur)
Predation Group Larry Madin (Chair), Erich Horgan (Rapporteur)
Vital Rates/Population Distribution and Dynamics Group Ted Durbin & Dian Gifford (Chairs), Charles Miller (Rapporteur)

Inter-annual variability

David Mountain (Rapporteur)
Slope Water Intrusions Jim Irish (Rapporteur)
Stratification Ken Brink (Rapporteur)

In the afternoon, a mini-lecture series was begun (the first of four) which was titled "Setting the context - Modeling activities in the Georges Bank Program". Dan Lynch presented an overview of the modeling activities in the program.

The goals of the modeling activity are to produce and make generally available the following:

The status quo as of the end of Phase I includes separate emphases on the climatological circulation on real topography and on idealized process studies. Larval fish modeling has been used to explore the biophysical linkages among the circulation, trophodynamics, and turbulent mixing. Most coupled studies have been in the context of climatological mean conditions; there has also been exploration of coupled response to wind events and to inter-annual variability.

During Phase II, projects funded will emphasize the following:

The Spatially Explicit Ecosystem Model which is emerging has the following character. There will be a detailed Eulerian description, on realistic topography, of the following:

Individual-Based Models for Larval Fish will be transported through this simulated biophysical environment; these models will integrate Lagrangian life histories for individual organisms in terms of:

Dan Lynch's overview was followed by short talks by:

Chris Naimie Modeling focused on the climatological environment.
Dave Greenberg A regional modeling perspective involving Hamilton Bank off Labrador to southwest of Georges Bank and including the Gulf of Maine.
Francisco Werner Coupled physical-fish larvae modeling.
Charles Hannah Supply pathways of particles onto Georges Bank.
Dan Lynch Calanus finmarchicus population dynamics.
Dennis McGillicuddy Simulation of Pseudocalanus population dynamics on Georges Bank.
Changsheng Chen 2D modeling of coupled physical and biology (NPZ) structure on Georges Bank.
Glenn Flierl Empirical Orthogonal Function (EOF) simplification of populations dynamics.
Mohammad Iskanderani Modeling the far-field boundaries - coupling the coastal model with a spectral element Basin Model.
Craig Lewis Affects of wind events on Georges Bank.

After a short plenary session at the beginning of Day 3 (Thursday), a mix of old and new groups again set to work. The PO Working Group focused on Physics modeling with Bob Beardsley, Chairman, and Charlie Flagg, Rapporteur, for the morning session, and Peter Smith, Chairman, and Bob Houghton, Rapporteur, for the afternoon session. The two biological groups continued to meet as well. The afternoon mini-lecture was titled "Biological structure of the target species and their rate processes on the Bank during 1995". Ted Durbin presented an overview of the distribution and dynamics of Calanus finmarchicus on the Bank in 1995 followed by a series of short talks on zooplankton by:

Bob Campbell Growth rates of Calanus.
Melissa Wagner RNA/DNA ratios in Calanus.
Jeff Runge Egg production rates of Calanus in the 1995 broad-scale time series.
Charlie Miller Calanus tooth development as an indicator of growth dynamics on the Bank.
Ann Bucklin Pattern of genetic variability in Calanus in the western north Atlantic as compared to other planktonic species.
Larry Madin 1995 target species predator fields on Georges Bank.
Bill Michaels The 1995 COP predator studies on Georges Bank.
Lew Incze The larval fish prey fields in relationship to small-scale turbulent fields.
Scott Gallager Feeding behavior of larval cod.
Elaine Caldarone Year to year variation in the RNA/DNA ratios of larval cod and haddock.
Cabell Davis VPR images of a section through a shelf/Slope Water entrainment feature caused by a warm-core ring.
Carin Ashjian VPR images of the physical and biological cross-bank structure.

Short reports by the group leaders from the previous days sessions were presented during the morning plenary session on Day 4 (Friday). These were followed by the continued meeting of some of the current groups and several of the interdisciplinary groups. These were: a morning working group on modeling the fish which was chaired by Dan Lynch with Chris Naimie as Rapporteur; an afternoon working group on mapping broad-scale distributions taking into account the flow field which Dan Lynch chaired with Chris Naimie as Rapporteur; an afternoon working Group on Drifters which was chaired by Dick Limeburner with Charles Hannah as Rapporteur.

The afternoon mini-lecture series consisted of presentations on various topics on the Physical Oceanography of Georges Bank by:

Charles Hannah Large Scale Inter-decadal variability.
Bob Houghton
David Mountain
Gulf-scale inter-annual variability.
Bob Beardsley 1994-95 Georges Bank southern flank stratification.
Jim Irish
Jim Manning
Jim Churchill
Slope Water intrusions and Scotian Shelf crossovers.
Dick Limeburner
Chris Naimie
Drifter observations and data/model comparisons.
Dan Lynch Data assimilation and hind-casting.

This session was Rapporteured by Glenn Flierl and summaries of the talks are given later in this report.

On Day 5 (Saturday), after brief working group reports at a plenary session, the morning working group focused on modeling the zooplankton. Ted Durbin and Dan Lynch co-chaired the session and Cisco Werner was the Rapporteur. The afternoon was spent by individuals working in small informal groups.

The principal activity on the morning and afternoon of Day 6 (Sunday) was the reviewing of plans for the 1997 field program. This started as a plenary session chaired by Peter Wiebe in which the cruise schedule was presented and Chief Scientists for the broad-scale cruises were selected (see Table 2 in the introduction to Section IV). The Chief scientists of the process cruises presented their plans for each cruise. A summary of the plans for the cruises are provided in Section IV below. The afternoon was devoted to smaller groups working out the details of the process cruises and the coordination of joint ship operations. In order to provide a complete picture of the plans for the cruises now scheduled to take place, included in the description of the cruise plans presented below is a description of a broad-scale cruise meeting that took place a couple of weeks after the workshop.

There were two general sessions on Day 7 (Monday). In the morning there was an open discussion of ways to improve the JGOFS data and information base structure and its usefulness. This produced a number of suggestions that are described in more detail below. In the afternoon, there was a general discussion of the planning required for Phase III of the U.S. GLOBEC Georges Bank Program.

Finally, on the morning of Day 8 (Tuesday), the election results were presented to the workshop participants. The election slate was finalized on the evening of 7 November. Voting began on the morning of 8 November (via a ballot and email) and ceased at noon on November 11. As of November 12, 1996 the new executive committee has the following members:

Bob Beardsley Woods Hole Oceanographic Institution
Jim Bisagni National Marine Fisheries Service, Narragansett
Ted Durbin University of Rhode Island
Scott Gallager Woods Hole Oceanographic Institution
Greg Lough National Marine Fisheries Service, Woods Hole
Dan Lynch Dartmouth College
Larry Madin Woods Hole Oceanographic Institution
David Mountain National Marine Fisheries Service, Woods Hole
Peter Wiebe Woods Hole Oceanographic Institution
Karen Wishner
University of Rhode Island

At the EXCO meeting after the workshop held on 6 December, Peter Wiebe was elected Chairman and David Mountain was elected Rapporteur.

Computers were very much in evidence during the workshop. Participants were encouraged to bring their computers with them to the workshop to aid in data sharing and synthesis. All computers could easily connect to the internet via 10 Mb/second, 10BaseT connections and 33 PC's, Macintosh, and Unix-based computers were brought to the workshop, with 28 of them connected to the Internet.

In addition to accessing the U.S. GLOBEC Georges Bank on-line database via the Web, participants were able to see several on-line data access and display demonstrations, access a local laser and color printer, use resources from their home organizations (e.g. using ftp and X-Windows), and keep in touch via email. It was not unusual to see investigators working on their computers until the conference room facilities closed down for the night at 2300.

This ready access to on-line data and information, provided by the large number of computers, facilitated the exchange of data and ideas; and fostered the opportunities for collaboration and data synthesis.

III. Working Group Reports

The following is a summary of the working group activities prepared by the Rapporteurs of each working group. First are the disciplinary working group reports followed by the interdisciplinary working group reports.

Physical Oceanography General Session

Wednesday morning November 6.
Robert Beardsley (Chairman), David Mountain (Rapporteur)

The physical oceanographers met as a group for the first morning session. Each PI or group of PI's made a short presentation to indicate areas of interest/objectives in the program, data collected and its status, and selected interesting findings. A table of the existing physical data and its current status was compiled during the workshop (Appendix III).


Broad-scale hydrography (David Mountain and Maureen Taylor): The objectives of the sampling are to provide a historic context for hydrographic conditions observed during the program period, to identify inter-annual variability, and to investigate event scale intrusions of water onto the Bank - most notably from the Scotian Shelf and the Slope Water region south of the Bank. Hydrographic data have been collected on six survey cruises in 1995 and six in 1996. The data processing is complete and the data are available through the GLOBEC Data Management System.

Noteworthy observations in 1995 were a large cross-over event of Scotian Shelf water in March and a Slope Water intrusion over much of the southern flank of the Bank in May. No similar intrusions were seen in 1996. The salinity of the waters in the region were considerably lower throughout the 1996 season in comparison to 1995.

Oxygen isotope data (Bob Houghton): Water samples for oxygen isotope (18O) have been collected on both broad-scale surveys and cruises by the Bedford Institute of Oceanography (BIO) as part of its mooring deployment program. On a 18O-Salinity diagram, the waters in Northeast Channel appear along a mixing line between end members representing a mixture of Labrador Shelf and St. Lawrence estuary water at one end and Slope Water at the other. The influence of Maine river input is indicated by a deviation from this line. In the central bank, the salinities were lower in 1996 relative to 1995, with Maine river influence evident in 1995 (and in data from earlier years), but not in 1996. The Scotian Shelf source water seemed similar in salinity in 1995 and 1996.

ADCP data (Charles Flagg and Julio Candela): Shipboard ADCP data have been collected on 37 cruises, from 1994-1996. All have been processed. Data from narrow band systems have been processed at Brookhaven, while that from broad band systems has been done at WHOI. The backscatter from the narrow band systems has been processed and work is continuing on the backscatter-to-zooplankton biomass relationship. This relationship is being developed using net data from the broad-scale surveys collected and processed by Ted Durbin. The data and cruise tracks are available on the Charlie Flagg's Stonybrook website. The URL is:

Julio is working on a better scheme for de-tiding the velocity data. While available at WHOI, they are not ready for final release. De-tided views of the data will be developed and made available. Work is also progressing on putting the narrow band data into an NODC ADCP format to make it easily available.

Satellite data (Jim Bisagni): Satellite derived (AVHRR) SST data from late 1993 to the present is on line and available. Also, gridded, optimally-interpolated SST maps are available - having been developed for a Gulf of Maine Regional Marine Research Program.

An interesting feature observed during 1995 is a cyclonic eddy located in Northeast Channel. A manuscript with Peter Smith was presented at the recent ICES meeting. The eddy was clearly seen in satellite data during April, moving to its inward most location (toward the Gulf of Maine) on May 4, and then moving back outward until May 10, when the SST signature was lost. It was also observed in Peter Smith's Northeast Channel east mooring, first at 100 m depth and then a few days later at 23 m depth. It was obvious in the temperature, salinity and velocity data sets. [Ken Brink pointed out that the tilt in the vertical axis could be used to indicate growth or decay of the eddy]. To suggest what was causing the movement of the eddy, the barotropic forcing of Ramp et al., (1985) for Northeast Channel flow was considered. In a frequency dependent, multiple regression analysis, the along-channel current did appear to respond to the cross channel wind, consistent with a wind induced setup/set down of the Gulf, as proposed by Ramp et. al., (1985).

Potential energy maps have been developed from the broad-scale hydrographic data. High stratification values associated with the Scotian Shelf water influx during February 1995 on the northeast peak of Georges Bank were comparable to the June values associated with seasonal stratification resulting from vernal warming.

Long Term observations

Drifters (Dick Limeburner): Eighty-three drifters have been deployed in 1995 and 1996. Seventy-five were drogued at 10 m depth and 8 at 40 m. Movies of the drift tracks are on the GLOBEC web site. In 1995, deployments were in the central bank region, while in 1996 deployment sites in northern Great South Channel (NGSC) and Browns Bank (BB) were added. Source regions for the Bank appear to be in the Gulf of Maine and NGSC - and to a much lesser extent, Browns Bank. Estimates of the mean flow in boxed areas have been made. In the central part of the Bank, no meaningful wintertime mean flow existed.

In 1995 a major event in late January took 5 buoys off the southern flank of the Bank. After that, the buoys largely stayed in the central region less than 60 m depth for a number of months - but with little indication of a strong re-circulation around the Bank. In summer, a racetrack-like circulation developed around the Bank, with some escapement into the Gulf of Maine - although those buoys largely came back onto the Bank. The loss of buoys across the southern flank was generally associated with rings.

In August 1996, a hurricane passed over the Bank and a large displacement was observed for drifter buoys in NGSC and on Nantucket Shoals - but not for buoys on the Bank itself.

Looking at the broad-scale hydrography, locally near the tidal mixing front on the southern flank the cross-bank density gradient changed sign from on-bank in winter (denser water in the shallow, central part of the bank) to off-bank in summer.

Moored observations (Peter Smith): A series of moorings are monitoring the inflows to the Gulf of Maine system through Northeast Channel (NEC) and around Cape Sable - from 1993 to the present. Cruises are also making ADCP and hydrographic measurements across NEC and around the Browns Bank-Cape Sable region. One mooring is located off Cape Sable and two mooring are in NEC - on the east and west sides of the channel (NEC-E and NEC-W). NEC-W was moved onto the northeast peak of Georges Bank from June 1994 to June 1995, to support the stratification experiment.

In the hydrography, in October 1993, an episode of high salinity (>35.5 PSU) and temperature (>15o C) in NEC deep waters (>100 m) was associated with a ring offshore.

Repeated ADCP sections were made across NEC to develop a de-tided velocity section showing the inflow on the east side and outflow on the west side. Net transports into the Gulf are consistent with earlier measurements (Ramp et al., 1985).

The focus of analysis is:

Long term moorings 1 (Jim Irish): The passage of hurricane Edouard on Sept 3, 1996 over the southern flank mooring caused a strong current response (order of 2 knots) with a baroclinic structure that was observed off-bank.

One should consider occurrence of internal waves-solitons on the southern flank as a possible retention mechanism. During year 2, the southern flank mooring had SEACATs at 20 and 30 m depth, sampling at 2 minute intervals. Large solitons seen in the temperature and density records during high stratified times were timed with the switch in the tidal current to the on-bank direction.

Long Term Moorings 2 (Bill Williams): Four 6-month deployments have been completed in the long term mooring program and the fifth period has been started. Observations have been acquired from October 1994 to October 1996. Three sites have been instrumented - the crest of the Bank (Oct 1994 - Sept 1995), the Northeast Peak (in the first year by BIO, but since then by this program) and the southern flank. On the southern flank, meteorological (MET) measurements have been made, although some parameters are missing for the first year of observations. Air temperature and PAR have been made at each site on each deployment. A full suite of MET measurements were made on the stratification mooring ST1 in year 1.

Stratification Study

Process mooring measurements 1 (Bob Beardsley): Surface MET data at the ST1 site have been used to estimate the surface heat flux (Qn) and wind stress. Initially Qn was negative during February and then became positive during the spring and summer.

The vertical thermal stratification appears to begin after the major Slope Water influx event in May 1995. The stability associated with the influx could have trapped the surface heating in a near surface layer, and thus triggered the subsequent development of stratification. The records show the influx of Scotian Shelf water in March and Slope Water in both May and August.

Process mooring measurements 2 (Steve Lentz): A depth-averaged temperature time series was calculated at the ST1 site and compared to the integral of surface heat flux Qn. The two curves agree on the seasonal time scale, suggesting that, to lowest order, a one-dimensional heat balance exists along the southern flank. More work is needed to check this initial description.

Retrospective heat flux calculations (Jim Bisagni): An equilibrium temperature heat flux method is being used to develop a historical surface heat flux time series for the Bank. Using model wind data from Fleet Numerical Oceanographic Center, shortwave data from Dick Payne at WHOI, dew point temperature climatology and satellite-derived SST, a computed heat flux series was used to compare with the measurements from ST1. The comparison is moderately good, and suggests that the method can be used to develop a historical heat-flux series for retrospective analysis of stratification development on Georges Bank.

BASS experiments (Sandra Werner - presented by Bob Beardsley):The BASS tripods were deployed twice at the ST1 site in 1995 - once when the water column was well mixed and once when well-stratified. Analysis will focus on the importance of stratification on the vertical flow structure and on bottom stress on a tidal time scale, and on developing a subinertial momentum balance.

Ring effects on the Bank (Jim Churchill): The response of Georges Bank currents and larvae to the passage of rings was examined. Three rings were considered during 1995. Ring A, in March-April, did not seem to influence the Bank. Ring B, in May, had a major effect, with the along-bank current switching from eastward to westward in advance of the approaching ring (that the currents were eastward was contrary to expectations of the mean flow - and both the ring and winds may have played a role in the occurrence of eastward flow). A surface buoy went westward at about 50 cm/sec in what appeared to be a frontal jet. In the moored data at ST1 the influence of Ring B was most evident at depth. The cross-bank residual current at ST1 can account for the progression of the Slope Water intrusion toward site ST2. Ring C passed by in August. Its influence was observed mostly in the near surface layer.

Visualizing model runs and observations (Jim Manning): a variety of visualizations of data were shown illustrating the use of different tools (e.g., AVS and MATLAB) to compare model runs and observations.

Micro-structure (Russ Burgett with Dave Hebert and Neil Oakey): The objectives of the micro-structure measurements are: to determine the processes and dynamics responsible for the vertical structure of the mixing rates; to determine the bottom stress at each location and a parameterization to explain the variability of the stress; and to determine the relationship between the biological patchiness and turbulent mixing rates. Measurements were made on two cruises - at two sites on each cruise, a shallow and a deep location. The first cruise was in late April-early May, when the water column was well mixed. The second cruise was in June when the water column was stratified. Micro-structure measurements were made using the EPSONDE profiler. ADCP and CTD measurements were also made. On the June cruise, Lew Incze made plankton pump measurements and Mark Berman and Jack Green made acoustic measurements with a TAPS system.

The dissipation time-series profile shows high values near the bottom and an obvious tidal signal. The velocity cycle leads the dissipation signal, with the lag increasing with distance above the bottom to maximum lags of over 3 hours. Bottom stress calculations are being done.

PO Inter-annual Variability Working Group

Wednesday afternoon, November 6.
David Mountain (Rapporteur)

The inter-annual variability evident in the broad-scale hydrographic data was discussed. The two major features identified were that the salinities observed in 1996 were about half a salinity unit lower than observed in 1995, and that the oxygen isotope analysis indicated no coastal Gulf of Maine contribution to the central-bank waters in 1996 (while the 1995 and earlier MARMAP data regularly showed a coastal Gulf of Maine contribution). At least two questions arise - what caused the change in salinity (a source water issue?) and why was there no Maine coastal contribution in 1996 (a circulation issue?)?

A number of steps were identified to investigate these questions:

As a separate item, the desirability of using the hydrographic data from the program to develop a climatology for the Bank was discussed. This analysis would be similar to the work of Charlie Flagg in the Georges Bank book, but it would be for the GLOBEC observational period. Regions of the Bank were identified, based on the broad-scale station grid, and the seasonal change in water properties in these regions will be determined. David Mountain will develop the seasonal temperature and salinity climatology for the Bank during the GLOBEC field years, while Peter Smith will compute the historical climatology from the AFAP data base using the same subdivisions of the Bank for reference. Bob Houghton will attempt to characterize the oxygen isotope versus salinity relationships for the same subdivisions from the GLOBEC and MARMAP data.

A review of hydrographic data from the NMFS spring and fall trawl surveys showed that the decrease in salinity noted on the Bank between the spring of 1995 and 1996 was evident in the Gulf of Maine as early as the previous fall (the fall of 1994 had higher salinities in the Jordan Basin region than did the fall of 1995).

PO Slope Water Intrusions Working Group

Wednesday afternoon, November 6.
Jim Irish (Rapporteur)

The group started out looking at the moored time series from the Stratification experiment which included the main Stratification Mooring (ST1) and the Long-Term Southern Flank Mooring in 1995. Jim Churchill had identified a few warm-salty events in the main Stratification Mooring data and related them to satellite images of warm-core rings. Jim Irish pointed out the same signals in the long-term moored data which existed before and after the main stratification mooring and included the fall of 1994. Based on plots from these two moorings, a number of significant warm-salty events were identified which appeared related to the presence of warm-core rings offshore. A ring was also identified which did not have any accompanying signature at the moorings. A number of satellite images and CTD sections were also studied to help interpret these events. Jim Manning and Dave Mountain pointed out a May 1994 ALBATROSS cruise which also appeared to document warm-core ring effects on the south flank of Georges Bank. This is interesting because we thought that the historical data did not show the effects of the warm water extending as far on the Bank as observed.

Besides the warm-salty events in the fall and summer, during the winter of 1995 several cold-fresh intrusions were observed in the moored observations which were attributed to Scotian Shelf water. Jim Bisagni's satellite images (between cloudy periods) also showed this cold water coming across the Northeast Channel.

Based on the identified warm-core ring and Scotian Shelf water events, a compilation was produced (Figure 1) which summarizes:

The results were further discussed, summarized, and presented in the late Friday afternoon physical oceanographic lecture.

PO Stratification Working Group

Wednesday afternoon, November 6
Ken Brink (Rapporteur)

The subgroup on Stratification, Mixing and Wind Driving met to discuss analysis plans and to arrange for cooperation and data sharing. The main topics discussed included:

PO Modeling I Group

Thursday Morning, November 7.

Bob Beardsley (Chairman) Charlie Flagg (Rapporteur)

Dave Greenberg described some results from a finite element model of the time dependence within the Gulf of Maine (GOM), Georges Bank and Scotian shelf. The model used a truncated Bay of Fundy with tidal boundary conditions to mimic the overall GOM tidal response. Results showed that the wind response reached near steady conditions after three to four days so that a steady model would be useful for many studies of the Bank. The along-shore wind stress produced an Ekman response over Georges Bank and Scotian shelf with a compensatory inflow through the NE Channel. Upwelling was evident within the Gulf and around the Bank wherever there was substantial topographic variation.

Changsheng Chen described a two-dimensional cross-bank model with which he investigated the separate effects of surface buoyancy flux and tidal and/or wind generated mixing. Heating alone from a winter-time initial condition led to a slight increase in anticyclonic flow around the Bank while cooling during the subsequent fall/winter decreased the anticyclonic flow and actually reversed the flow on the southern flank. Surface heating with tidally generated mixing resulted in a greater increase in anticyclonic flow during the summer than heating alone. After cooling reduced the anticyclonic flow during the fall/winter, the second heating cycle produced intensified anticyclonic jets along both the north and south flanks. There was a discussion about whether the density gradients between the crest and southern flank matched those shown by Dick Limeburner.

Frank Bub demonstrated some results from his 12 regions by four depth bin box model for the Gulf of Maine/Georges Bank system. Adjustable variables used to match observed temperature/salinity distributions were transports and the vertical and horizontal turbulent exchanges between the boxes. The model was forced with inflow from the Scotian shelf and required to match mean conditions for each box derived from the AFAP data base. The model showed a cross NE Channel flow reminiscent of observations. Discussion centered on the lack of a deep Slope Water inflow through the Channel. It was suggested that the necessary inflow was supplied from upper slope boxes on either side of the Channel.

Wendell Brown described an analysis of coastal sea level and bottom pressure records within the Gulf of Maine showing a vigorous response due to wind with 8 to 13 day periods. Maximum sea level response was due to nearly alongshore winds from the east or west. The maximum pressure difference response, proportional to geostrophic transports, occurred at 3 to 4 day periods. Based upon the results from Greenberg's frequency response results for the GOM, comparisons were then made to daily steady-state model results driven by "Halifax" sea level and 36-hour-averaged winds. Although there was general agreement, the steady-state model under estimated the sea level response for strong events.

The last presentation was by Chris Naimie in which he indicated that 4D results from the Dartmouth Finite Element Model, based upon climatological forcing were archived and available to observationalists for comparison with their data. Comparisons with moored data will be available on the Dartmouth WWW site. Chris then described the modeled asymmetrical response to tidal forcing on the northeastern flank of the Bank. Due to the bottom slope and density stratification, onshore near bottom flow produced increased stability over the upper flank while down-slope flow resulted in reduced stability and increased mixing. Discussion centered on how the model dealt with turbulence. It was noted that the hydrostatic condition and turbulence closure assumptions precluded the model's ability to explicitly include high frequency phenomena such as tidally generated solitons and the localized turbulent mixing that they produce.

PO Modeling II Group

Thursday afternoon, November 7.
Peter Smith (Chairman) Bob Houghton (Rapporteur)

The session began with a discussion of whether or to what extent model calculations of features such as Lagrangian drifter tracks compare with observations. The success is mixed. The model is a representative of climatology and not the response to forcing of a particular year. There was discussion of the model response and robustness to stochastic forcing of extreme events such as storms or Gulf Stream rings. Chris Naimie, Charles Hannah, and Dick Limeburner were going to pursue the issue of whether model runs could be useful in designing drifter deployments in Phase II.

Dan Lynch described plans to run hind-casts of the response to observed 1995 and 1996 forcing. The domain would be confined to the Georges Bank region covered by the broad-scale sampling (~150 m isobath) and boundary conditions would convey all information of the far-field forces and water properties. These boundary conditions will be generated in part by matching the observed velocity field.

Model calculations will be used to identify the role of Gulf Stream rings in generating the Slope Water intrusions. Except for some of the process study sections, little is known of the offshore fields advected onto the outer Bank. The May 1995 intrusion was best studied, especially by biologists, and should be examined first.

A description of the inter-annual variation is not yet fully defined, but it is hoped that the hind-cast calculation will shed some light on 1995 and 1996 differences.

For the stratification, it is thought necessary to first describe the field data and get the forcing right, and then hind-cast. An important question concerns the connection between vernal stratification and Slope Water intrusions "do they affect the precondition?" Model runs with and without heat and tides would be useful.

The Drifter Group

Friday afternoon, November 8.
Dick Limeburner (Chairman) Charles Hannah (Rapporteur)

The drifter group examined the GLOBEC drifter trajectories and simulated drifter patterns (from the climatological seasonal mean fields). The emphasis was on developing descriptions of the flow patterns and understanding the utility of the seasonal mean flow.


We explored the idea that the seasonal cycle can be described by the transition between two states.

Further work is planned to provide a quantitative examination of 1) the seasonal spin-up and spin-down of the re-circulation time scale; and 2) the seasonal cycle in near surface velocity at key locations around the Bank. A publication is planned.

Dick Limeburner led a lively discussion of the placement of the drifters during 1997 which will focus on the pathways from the Gulf of Maine to Georges Bank. The tentative deployment plan is shown in Figure 2: 'Satellite tracked drifter deployment sites for the 1997 cruises.'. The top panel shows where drifters would be deployed in January and February, the lower panel for the rest of the spring and summer.

Summaries of Friday afternoon mini-lectures, Glenn Flierl (Rapporteur)

Decadal-scale Variability (Charles Hannah): From historical hydrographic data, the 1958-1968 period was 1o to 2o C colder while the 1972-1987 period was 1-2o C warmer than average. The bottom temperatures show similar changes, though the magnitude was about 1o C with a maximum of 4o C along the shelf break. During the cold years, the geostrophic transport was about 2 Sv higher relative to 1000 m. The circulation on the Bank did not change strongly.

Seasonal Variability (Bob Houghton): The 18O/S relationship is much tighter than the T/S relationship and can be used to distinguish water masses and the fractions of these in a particular sample. On the broad scale, one can identify Labrador Shelf Water, St. Lawrence River Estuary Water, Scotian Shelf Water, and Labrador Sea Water. The NE channel had a mixture of Labrador Shelf and St. Lawrence waters. Salinity on the Bank showed 0.8 PSU variation with 1996 being 0.3 to 0.5 PSU fresher. This variation occurred over the full depth of the Gulf of Maine. In 1995, Georges Bank was not on the mixing curve, implying about a 20% contribution from GOM coastal waters. In 1996, there was no GOM coastal water in the Bank samples. Perhaps the spring fresh water was too light to get onto the Bank. There were significant changes in the properties of the source waters.

Stratification (Bob Beardsley): Observations of the stratification on the southern flank showed that it was reduced by mid-November, presumably from cooling and storm-induced mixing. A Slope Water intrusion was seen in early November and a Scotian Shelf intrusion in mid-December. The surface waters warmed by mid-March. Scotian Shelf and Slope Water intrusions were seen in May.

Intrusions (Jim Irish): The Slope Water intrusions were characteristically warm and salty, while the Scotian Shelf water was cool and fresh. Seven events were observed during the period May 1994 to September 1995. The changes in salinity and temperature were as big as the annual signal. A streamer from the Scotian Shelf could be seen by satellite during one intrusion. The signal was detected at various depths and often looked like a sloping front advecting across the mooring. During May, there was onshore to offshore movement.

Drifters (Dick Limeburner): In 1995, drifters were seeded at 5 sites; in 1996, various other source sites were used with some floats being placed on Browns Bank. The floats were drogued at 10 m, but did move relative to the water and diverged from true fluid trajectories over a few days. January-October 1996: at the end of January, 5 tracks left the Bank via a Ring. July-October: an organized circulation pattern was visible. Most of the water drawn off the Bank went out to Rings. In 1996, there was an anticyclone in the Great South Channel entrance. Wind events were visible and affected the flow most strongly within the 60 m isobath. The mean flows were weak over the winter; circulation developed on the north flank in July. During July and August, there was reverse flow on the NW side. After July, there was northward flow in the Great South Channel (which also appeared in the model calculations). During the 1996 winter, there was SE drift. It appeared as though the strong jets were unstable and ejected eddies which may have caused significant mixing from the GOM. Again, drifters crossed the shelf-break due to Rings.

Data Assimilation (Dan Lynch): The Dartmouth team will be producing detailed 3-D circulation maps for individual cruises, including both broad-scale and process cruises. Early attention will be on 1995. These will be in the form of a hind-cast of the physical environment, obtained by assimilating moored, CTD and ADCP field data. The geographic scope of the hind-cast will be approximately the same as the broad-scale survey, i.e. it will cover the bank per se, roughly shoalward of the 150 m isobath. Resolution will be approximately 1 km horizontally. Each hind-cast will have estimates of error and will be archived along with the cruise data. Because the assimilative model is tide-resolving, there is no need to de-tide the ADCP data prior to its assimilation. Drifter trajectories will not be assimilated; they will be used as a check on the Lagrangian impact of the computed error/uncertainty fields.

Vital Rates/Population Distribution and Dynamics Group

Wednesday and Thursday mornings, November 6 & 7. Ted Durbin & Dian Gifford (Chairmen), Charlie Miller (Rapporteur).

It was decided to have each investigator characterize information available or developing on the target species for which they are responsible. Notes on those talks are ordered here by group. The discussions overall gave a sense of steady progress with investigations of both fish and copepod target species.

Fish Larvae

Broad-scale Survey: Abundance and Otolith Analysis: For the 1994 process cruises, single broad-scale cruise and COP cruises, larval fish were sorted initially at sea. More intensive resorting of the process cruise samples has been completed in the laboratory at NMFS (G. Lough). No additional sorting of the broad-scale or COP samples is currently planned. Jack Green reported that sorting of the 1995 broad-scale larval samples is underway at the Sorting Center in Szczecin, Poland. Data for the May 1995 cruise are due from the Center in December. All 1995 samples will be completed under the current agreement. High priority has been assigned to March, April, and May, with second priority to February, June, and July. No funding as yet been provided for analysis of 1996 samples, although some 1996 bongo tows have been included in the work contracted for 1995 materials. A "sort" at the center includes removing all fish larvae, identifying them and both measuring and storing cod and haddock larvae in vials of ethanol. Vials will be returned to J. Green for analysis of birth date by otolith ring counting. It is not anticipated to study otolith ring widths to estimate growth rates as a function of time. The location and responsibility for larval fish studies in the GLOBEC Georges Bank Program will be changing soon. Sandy Hook Laboratory will be phasing out of the cod-haddock work after working up the 1996 data, and all of the GLOBEC work will transfer to the Narragansett NMFS facility for the 1997 sampling and beyond. Funding for sorting the 1997 and 1998 broad-scale samples is in the budget approved by the program.

Physiological Condition and Growth: Elaine Caldarone demonstrated a strong, positive correlation from experimental data between the ratio RNA/DNA and the rate of increase of protein (growth) in larval cod. Most of the field estimates of RNA/DNA for larval cod and haddock are complete through 1995. Years studied to date can be rated as follows:

Fish from station 17 of the April 1995 broad-scale cruise were in very poor condition compared to fish sampled from the rest of the Bank.

Larval Fish Ecology: A full picture of studies of larval fish ecology carried out during Process Cruises was not available to us because Greg Lough was not able to attend the workshop.

Cod Larval Feeding Studies: Elaborate experiments on feeding by early stage cod larvae were reported by Scott Gallager. Larvae newly hatched from laboratory-maintained cod are taken to sea and allowed to feed on natural foods. The foods (protozoa) are size fractionated, labeled according to size with fluorescent dyes varying in emitted color and recombined. Larvae are allowed to feed, then collected. Gut content analysis by epifluorescent microscopy shows which sizes of food are consumed. Very early larvae eat protozoa as soon as the mouth is open and up to at least 7 days of age, well before yolk sac absorption is complete. Crustacean nauplii are not eaten in this interval. Upcoming studies in S. Gallager's laboratory focus on the role of the light field in visual spotting of prey by larval cod. It is already clear that most prey are taken from below. In tanks they are silhouetted against the apparent darkness outside Snell's window. How this operates in the field with a chaotically oscillating water surface will require modeling work.

Zooplankton Ecology

Broad-scale Survey Studies of Target Species Abundance: Ted Durbin showed stage proportions for Calanus finmarchicus from selected stations from all cruises of the 1995 broad-scale series. Different stations seemed to start accumulating a stock of younger Calanus stages at different times ranging from January through March. Most stations showed two generations, one with prolonged egg input and slow development before May, the other after. There was sharp hiatus in both naupliar and copepodite abundance during May at all Bank and Gulf of Maine locations. Station 38 in the southwestern outlet of the Gulf of Maine showed a strong February to April generation, but no early stages after May. This one generation pattern was restricted to that region. T. Durbin reported that the planktoneers have counted 1-m2 MOCNESS samples from all priority 1 stations from the 1995 broad-scale cruises and from priority 1 and 2 stations for 1996. Data books are available for 1995. A process of conversion to a database format is underway. T. Durbin presented a series of questions that he intends to investigate using the data now available.

Recruitment Rates of Calanus finmarchicus: In addition to work by the URI plankton group, Jeff Runge has counted adult females of all species from all of the 1995 broad-scale samples. He showed maps developed from those data by "kriging", a two-dimensional method of data smoothing and contouring. From each station 30-50 female C. finmarchicus were also rated with an index of reproductive condition (scale of 1=unripe to 7=ready to spawn shortly). Using a correlation (r) between the index and actual egg production developed experimentally on process cruises (r2 = 0.7), the total expected egg production at a station can be predicted from the product of expected individual fecundity and estimated female abundance. Kriged maps of this variable across the Bank were mostly similar to the abundance maps, since most females had high indices at all places and times. Strong concentration of early spawning output occurred at Gulf of Maine stations and over the northeast peak. May productivity was less than that of earlier months, but still high enough to make the May hiatus in abundance of all naupliar stages surprising. By May, the near absence of females over the shoaler areas of Georges Bank made a large blank hole in the spawning output map for Calanus.

Calanus finmarchicus Growth Rates: Bob Campbell of the URI group showed molting and growth rates for copepodite stages of C. finmarchicus estimated from experiments on Process Cruises of 1994 and 1995. Animals were sorted by stage and a sub-sample weighed. They were held in natural seawater for a period, then staged and weighed again to determine molting rates and weight gain. Most values in a very broad scatter of results were positive with a vague central tendency around 0.1 d-1, which is a fairly dramatic number (biomass doubling times of only one week). A laboratory experiment at full nutrition was carried out during summer 1996 with which to compare these results. Stage-wise development time data from these experiments are of a new standard, showing progressively longer intervals in each stage and cleanly equi-proportional change in stage durations at different temperatures. These data will be found indispensable by the modeling group. Similar data developed by EU-TASC at the Bergen mesocosm facility will provide a good comparison for stocks from the eastern side of the range.

Calanus finmarchicus Development: Charlie Miller reported two studies on C. finmarchicus development in the field.

Calanus finmarchicus Feeding: Feeding experiments with field-collected C. finmarchicus in late stages were described by Dian Gifford. She stated that as the copepod development season progresses the copepods are more and more dependent upon heterotrophic protozoa as food. She found that April cruises show very low levels of food abundance in the 5 to 20 m range most favored by Calanus as a diet. This may match Miller's observation of an apparent hiatus in developmental progress sometime in early April 1996. D. Gifford noted that at times, the pelagic protozoan community is dominated by Mesodinium, believed to be a "mixotroph". Mesodinium is an ideal size to serve as copepod food, but it has astounding capability for making escape jumps which apparently protects it from predation.

Genetic Studies of Target Copepod Species: Genetic studies by Ann Bucklin have shown relatively low levels of genetic variation of a mitochondrial gene (16S) in C. finmarchicus, compared to levels found in Acartia for the same gene and Meganyctiphanes for two other mitochondrial genes, Co1 and Cytochrome B. She can differentiate west Atlantic from east Atlantic stocks of Calanus, but the predominant genotype was the same in all regions, and no distinctions were consistently found within regions. She is currently examining other population genetic characters for Calanus, intending to check conclusions based on mitochondrial genes. Planned efforts at Norwich in the UK will allow a trans-Atlantic comparison. A. Bucklin also reported that a PCR-based species identification system has been developed for Pseudocalanus species common on and around Georges Bank. This will be applied to determine the relative importance of P. moultoni and P. newmani at different times and places in the region.

Zooplankton and Particle Distrbutions from the VPR: Cabell Davis and Carin Ashjian reported that 10 Video Plankton Recorder transects have been fully analyzed from GLOBEC studies over Georges Bank. Horizontal and vertical zonation of abundant zooplankton (and in some cases phytoplankton) species are very well characterized in these data, at least for adult stages that are readily recognized. Some species are well characterized at several late development stages. C. Davis and C. Ashjian showed some of the data, which are too complex to characterize simply here. An enormous increase in data production rates is expected early next year when new hardware and software for image analysis come on line. This system recognizes actual species categories in images with about 95% accuracy. The system is slow compared to human operators, but it works tirelessly.

Small-scale Vertical Distribution of Zooplankton: Lew Incze reported the availability of pump profiles at 1 meter resolution from the sea surface to 50 meters at 16 stations collected in May 1995 in conjunction with Micro-structure studies by Dave Hebert and Neil Oakey. He hopes to begin counting of these profiles soon.

Predation Group Sessions

Wednesday and Thursday, November 6-7.
Larry Madin (Chairman), Erich Horgan (Rapporteur)

Data Status: Initial discussions began with review of data from the 1994 and 1995 field seasons and the status of sample analysis. Both hydrographic and zooplankton data (1-m2 MOCNESS and 10-m2 MOCNESS) are now available for all process cruises in 1994 (AL9403) and 1995 (SJ9503, SJ9505, SJ9507, EN268):

Formatting of MOCNESS data was discussed; the GSO and WHOI groups plan to coordinate with the Durbin Process Group to choose a similar format for entry of the data into the GLOBEC Data management system. Existing search and graphics tools in the GLOBEC database will probably be used for retrieval and analysis. It was agreed to generate a list of the remaining unanalyzed 1-m2 MOCNESS and 10-m2 MOCNESS hauls to understand where holes in field collections and disposition of the samples occur. Scarce support for additional sorting makes it important to prioritize remaining samples.

Predator Feeding Rates: Feeding rate data for the major predators that have been investigated so far were discussed. This is summarized below. If all stages of the target prey species (eggs through adult for copepods, all sizes of larval fish) are considered, then there are potentially more predators than have previously been considered, including omnivorous copepods and sand lance. Immunoprobes were suggested as a means to screen a large diversity of other possible predators for the presence of Calanus or Pseudocalanus. Immunological screening for Calanus is planned to begin in January 1997. Sand lance are difficult to sample, but Bill Michaels offered specimens from his collections for preliminary analysis, and there are feeding data for sand lance in the literature.

There was further discussion of immunological and molecular probes for detection of prey species. The immunoprobe for Calanus seems to be working well in preliminary tests, and Madin/Bollens have funding to support development of a similar polyclonal probe for Pseudocalanus spp. On Friday, Madin and Horgan visited the Bucklin and Kocher labs at UNH to discuss application of DNA probes for identification of target species in predator stomachs. Cod and haddock specific probes have been developed by Rosel and Kocher with COP funding, and these probes could presumably be available to the Predation Group for detection of larval fish as prey of invertebrate predators. Specific DNA probes for copepods developed by Ann Bucklin may be valuable for detection of the copepod target species. We plan to work with these groups to try the probes in comparison with the polyclonal methods already under development.

Summary of feeding data:

Provisional Estimates of Predation Mortality: After reviewing data on abundance and feeding rates of major predators, one cruise (May 1995, SJ9507) was picked and estimated predation intensity at the Crest and Flank sites was compared (Table 1). In this exercise, abundance data were used from process cruise tows and also from broad-scale tows (using 1-m2 MOCNESS data from Durbin group) along with the best estimates of feeding rates to calculate a range of predation effects on the prey groups. These are summarized in Table 1. This is the approach that will be used generally in combining rate data with distribution maps of prey and predator from broad-scale results. Predation rate functions need to be further refined to take temperature and prey abundance into account for the major predator species.

The preliminary exercise assumed a mixed water column, but there are plans to make future calculations with depth-stratified data. The estimated predation impacts were far greater at the Crest site (range: 2% to 112% prey stocks d-1) than at the Flank site (range: 0.07% to 6.5% prey stocks d-1). In both cases, predation impact was largely a function of predator abundance, measures of which are often highly variable between adjacent collections.

This method of estimating predation effects, and the sample results were discussed with representatives of the modeling groups (D. Lynch, C. Naimie, C. Davis, D. McGillicuddy, et al.) on Friday afternoon. A goal agreed upon by the predation group and modelers is to produce depth-stratified maps of predator-prey co-occurrence which can be used to estimate predation mortality for each prey category, depending on the predation rate functions which connect predator-prey pairings. Model development will then proceed to incorporate information from day and night differences in predator-prey distribution, which will begin to address the behavioral component of predation.

Barbara Sullivan and Carol Meise have already analyzed previous MARMAP data on predator abundance on the Bank, and these data (in forthcoming DSR volume) were recommended for use as an initial data set by the modelers.

Table 1.

Estimates of predation impact at Crest (mixed) and Flank (stratified) sites in May 1995. Prey abundance data from broad-scale 1-m2 MOCNESS data (Durbin's group). Ranges of predator abundance are from broad-scale and process 1-m2 MOCNESS data (Sullivan's group) to illustrate variability in predators and its effect on impact. Rate data from gut-contents analysis plus digestion times, or from lab feeding experiments, except Themisto and "shrimp" from literature data. Predation impact and percent stock removed d-1 assume no production of prey or removal by other means. All calculations assume a homogeneous vertical distribution of properties.

rate impact%

#m3 #m3
#m3 d-1
CREST Site MAY 1995
hydroids copepod eggs ? * 678 3.6 ml h -1
hydroids copepod eggs ? 13,000 3.6 ml h -1
hydroids 1000 nauplii 678 1.9 ml h -1 32 3%
hydroids 1000 nauplii 13,000 1.9 ml h -1 624 62%
chaetognaths 1875 copepods 44 3 copepods d-1 132 7%
chaetognaths 1875 copepods 165 3 copepods d-1 495 26%
Themisto gaudichaudi
1875 copepods 4 10 copepods d-1 40 2%
Themisto gaudichaudi
1875 copepods 30 10 copepods d-1 300 16%
"shrimp" 1875 copepods 4 10 copepods d-1 40 2%
"shrimp" 1875 copepods 0.06 700 copepods d-1 42 2%


1875 copepods 0.6 701 copepods d-1 420 22%
hydroids copepod eggs ? 15 3.6 ml h-1
hydroids copepod eggs ? 74 3.6 ml h-1
hydroids copepod eggs ? 450 3.6 ml h-1
hydroids 9100 nauplii 15 1.9 ml h-1 6 0.07%
hydroids 9100 nauplii 74 1.9 ml h-1 32 0.30%
hydroids 9100 nauplii 450 1.9 ml h-1 180 2%
chaetognaths 4000 copepods 33 3 copepods d-1 100 2.50%
chaetognaths 4000 copepods 60 3 copepods d-1 180 4.50%
Themisto gaudichaudi
4000 copepods 0.5 10 copepods d-1 5 0.10%
Themisto gaudichaudi
4000 copepods 7.2 10 copepods d-1 72 1.80%
"shrimp" 4000 copepods 8 10 copepods d-1 80 2%
"shrimp" 4000 copepods 26 10 copepods d-1 260 6.50%

* egg abundance data unavailable at time of Workshop

Fish/Trophodynamics Modeling Group

Friday morning, November 8.
Dan Lynch (Chairman), Chris Naimie (Rapporteur)

Cisco Werner began the discussion by describing the historical development and general capabilities of his individual-based trophodynamics model. The trophodynamics component simulates certain aspects of the early life history of larval fish on Georges Bank within the diagnostic seasonal circulation fields from the finite element model developed by Lynch and colleagues at Dartmouth College. The model traces the effects of feeding on growth of larvae and predicts survival based on individuals exceeding a certain age-dependent critical weight ("death barrier"). The prey fields for the larvae are frozen in time and compartmentalized in space with vertically uniform densities. The salient results of the model are:

Problems identified with the present version of the model include the fact that the actual prey distributions are not observed to be uniform, but rather aggregated at particular depths; some prey types are missing (e.g., microzooplankton); and that the effect of predators on the larvae needs to be included. Vertical migration of the larvae should be refined and some aspects of the foraging behavior and nutritional benefits of the prey species may be in error. During the discussion of the models a summary illustration of factors influencing the growth and abundance of larval cod on Georges Bank was created (Figure 3,).

To assess the predation question, it was suggested that a simpler "holistic" model could provide a complementary view of the problem in relation to the more mechanistic treatment of feeding in the trophodynamics model (Figure 4). The simpler model would treat larval growth rate as a known (exogenous) variable based on RNA/DNA field observations, and include several modes of mortality, including starvation, predation and advective losses. The larvae in this model would be allowed to migrate vertically and both predation mortality and growth rate fields could be treated in either a Lagrangian (drifting) or Eulerian (fixed) manner, with an additional stochastic component.

The session concluded with a wide ranging discussion of the details of the foraging behavior of larval fish. These aspects included:

Mapping Working Group

Friday afternoon, November 8.
Dan Lynch, (Chairman); Chris Naimie, (Rapporteur)

A number of scientists involved in the collection of data during the 1994 and 1995 broad-scale surveys (i.e. AL9404 ( 31 May - 10 June 1994), EN259(10-20 Feb 1995), EN263(13-24 Mar 1995), EN265(11-23 Apr 1995), AL9505(9-18 May 1995), AL9506(5-15 Jun 1995), AL9508(10-21 Jul 1995)) are interested in creating synoptic distributions of cruise data. Of concern in this regard are processes which effect the distributions at temporal scales equal to or shorter than the duration of a particular broad-scale cruise (typically of order 10 days, as listed above).

The "synoptic correction" could best be conducted using a hindcast of the coupled physical/biological system for each individual broad-scale cruise. Given that such simulations are not feasible at the present time, subsequent discussion centered on the value of performing "advective synoptic corrections" using the vertically averaged circulation from numerical simulations of the climatological bimonthly circulation recently conducted at Dartmouth College. The group generally agreed that comparison between distributions of data at actual observation positions and at synoptically corrected positions could provide additional insight under some circumstances. This optimism was based mainly on:

However, some concerns were raised regarding the appropriateness of making such corrections: