Project Summary

Ken Brink, James Irish, Robert Beardsley, Richard Limeburner

U.S. GLOBEC: Long-term ADCP, Moored and Lagrangian Measurements and
Analysis as part of the Georges Bank Study

I. Introduction:

II. Proposed Research:

A. Telemetry from 1999 Moorings (Irish, leader)

As part of the Long-Term Moored Program, both GOES and ARGOS telemetry are utilized to return data daily to WHOI. In the 1999 field year, Beardsley and co-workers are proposing separately a Northeast Channel crossover experiment where knowledge that water from the Scotian Shelf has reached Georges Bank would trigger a rapid response survey. Much of the software for this effort is in place, and setting up the system should require minimal effort. The real-time data will also be accessible to other GLOBEC investigators.

B. Analysis of Moored Array Data (Brink, Beardsley, leaders)

The funds currently requested will cover the cost of editing, processing and archiving the results from the last (1999) mooring deployment. This will lead to the availability (to GBS investigators) of a five-year moored array data set on the web which will be supplemented by relevant sea level, air pressure and wind records from around the Gulf of Maine and along the Scotian shelf. Some scientific analysis of the existing moored array data set (1994--1996) is presently underway. These efforts tend to involve special events, or other problems that do not require (or are not better done) with a five-year time series.

When 2000 arrives, we will have a rather complete, 4--5-year set of time series at the Northeast Peak and South Flank locations, along with supplemental data. At this point, analyses can begin that involve maximal statistical confidence or ability to compare among years. Broadly speaking, these analyses can be divided into physical and interdisciplinary approaches, but all analyses will contribute to the overall GBS goal.

The purely physical analyses can be summarized as follows.

• i) Wind and remotely alongshore forcing. We anticipate that the energetic variations on wind-driven time scales will be critical in terms of determining Lagrangian paths followed by individual particles. We will be able to relate, both statistically and by event analysis, how current fluctuations on the Bank (with time scales longer than about a day) relate to nearby winds and remote effects. The tools we plan to use are relatively standard (correlations, cross spectra, EOFs and frequency-domain EOFs, for starters), but the important opportunity is the duration and quality of the data set. Length is important since it allows, alternatively, high statistical confidence or the ability to resolve year-to-year changes in variability.

• ii) Intrusion events. The moored array has already detected some major intrusions of offshore waters onto the Bank. There is never a strong signature of saltier offshore waters being left behind. But, a weak signature does not mean an unimportant influence. We hypothesize that the onshore flux is associated with vertical turbulent mixing during onshore intrusions. This hypothesis will be difficult to test using moorings unless the alongshore advective time scale is slow relative to the intrusive time scale. Very likely, dealing with the intrusion questions will require making use of the extensive hydrographic and remote sensing data base available to us.

• iii) Seasonal patterns. Our five-year time series, along with historical information (e.g., Butman and Beardsley, 1987), should provide enough information to estimate, with some confidence, the seasonal patterns of mean flow, variance, stratification and other physical properties. Further, this means that anomalies from these cycles can be treated, and compared to both local and basin-scale forcing functions. The length of the available time series is rare in oceanography, especially for direct current measurements, and we intend to exploit this longevity.

The extensive moored optical data set provides an opportunity for interdisciplinary analysis. The transmissometers and fluorometers can provide estimates of phytoplankton standing stock. This stock time series, in turn, can be used to relate any surges of productivity to spring restratification, upwelling events, or interactions of Bank waters with warm core rings. Thus, there is an opportunity to use correlative approaches to estimate what physical phenomena affect primary productivity. In addition, the optical data provide a good basis for exercising bio-optical models.

Finally, we intend to use the five-year moored array data set to deal with the central GBS question: what physical effects act to control recruitment? By about 2000, we expect that our biological colleagues can produce numbers on annual, Bank-averaged recruitment of cod and haddock, as well as populations of the key zooplankton species. These five-point (one number per year) time series can then be used with the moored array data to deal with central hypotheses. For example, is recruitment affected by stratification, ring intrusions or storminess, and at what critical times? Admittedly, a five-year time series may be too short to answer questions such as these, but we can at least expect to be able to eliminate some hypotheses. We anticipate, further, that our comprehensive data set will also open the door to other sorts of opportunistic interdisciplinary cooperations.

C. The Drifter Program (Limeburner and Beardsley, leaders)

The present Phase II grant provides funding for the field effort described above through September 1999. We request here funding to start October 1999 (year 2) to (a) complete the collection of ARGOS data for those drifters still functioning in our study area, (b)complete basic processing of all drifter position and temperature data and associated surface meteorological and satellite SST data used in our drifter trajectory animations, (c)construct and disseminate animations of the combined data via the GBS web site, and (d)analyze and synthesize the combined data and prepare manuscripts summarizing these results.

We plan to focus initial analysis efforts on:

• i) The near-surface bank response to extreme wind events (e.g., during Jan/Feb, 1995; April, 1995; August, 1996; and April, 1997),which can cause major movement of drifters onto and from the Bank. The vector correlation between wind stress and Lagrangian current over Georges Bank is low in general, but Bank waters do seem to respond to winds above a critical strength in a manner that is downwind and strongly coherent (slab-like). The observed strong response to along-bank winds is consistent with recent idealized numerical model results. Further work is needed to compare the observed and model responses to cross-bank winds. We have recently obtained detailed digital bathymetry (horizontal resolution 460 m), which will allow us to determine water depth along each drifter's path and decompose the drifter current into along- and cross-isobath components for more detailed study. This will allow us to examine if small-scale finite-amplitude bottom topography over the crest has any direct influence on the subtidal drifter motion there.

• ii) A comparison of the 1995--1997 observed drifter trajectories with the bimonthly seasonal flow fields produced using the Dartmouth Circulation Model and climatological forcing. This work is a collaboration with C. Hannah and C. Naimie, who have already begun to deal with Lagrangian model-data comparisons. The approach is to divide the Bank into sections, and then ``deploy'' model drifters in each section at the same starting place and time as the real drifters. The statistics of the observed and model drifter motion within each section will then be then compared.

• iii) The transition between winter and summer flow regimes and the relationship to the cross-bank density field. As mentioned above, the cross-bank density gradient reverses sign during winter and again in late spring. The salt and thermal contributions to this reversal are about equal. The annual density variation on top of the Bank is equivalent to an 8-cm change in sea level at constant bottom pressure. The cross-bank density reversals occur in the tidal mixing front (over a distance of about 12~km, about two internal Rossby radii of deformation). We plan to analyze broad-scale CTD data collected near the tidal mixing front to characterize the density reversals statistically, and then present a conceptual model of the horizontal pressure gradients and geostrophic currents associated with the buoyancy-forced time varying pressure field.

• iv) The exchange of shelf/slope water due to warm-core rings. We have produced animations of SST images with overlaid drifter trajectories or area-mean currents to begin to get a sense of how often exchange of shelf/slope water due to rings takes place. We propose to continue this effort by working with other GLOBEC investigators (e.g., R. Schlitz, K. Wishner, C. Lee) to combine drifter, moored, and shipboard (CTD, ADCP, SeaSoar, and biological) data with well-resolved AVHRR and SeaWiFS imagery to produce estimates of the fluxes of water, heat, salt, nutrients, and fish larvae due to shelf/warm-core ring interactions.

We anticipate additional topics, especially related to the tidal mixing front, to arise as our analysis efforts intensify.

D. Underway ADCP Analysis (Flagg, leader)

Continued shipboard ADCP Data Collection and Processing:

• i) We will continue routine NB-ADCP data collection, which is currently funded through December 1999. We will need continuing support to complete the data processing and archiving not completed at that time. We also need support to accommodate the increasing variety of GBS research vessels, many of which are equipped with BB-ADCPs that present significant logistical difficulties in orchestrating a coherent ADCP measurement program. Routine ADCP data collection on 10 to 30 cruises per year requires a more energetic interaction with the ship operators and ship technicians than we have been able to support in the past.
• ii) We need to develop BB-ADCP processing and archiving methods to make these data readily available. We also will deal with the backlog of BB-ADCP data sets that have developed. This is especially important now that more GBS cruises will use only BB-ADCPs. We will expand the present CODAS data processing system to embrace the BB-ADCP data stream as easily as it now deals with the NB-ADCP data.

Shipboard ADCP Analysis Projects:

• i) We will update the tidal analysis based upon the ADCP data to encompass all the data collected to date. Since tides produce the vast majority of current variance, it is vital that their effect be removed as reliably as possible to allow an examination of the non-tidal flow field. Presently, we predict tidal currents for the central portion of the Bank based upon ADCP and moored data from 1994 through 1996 using the method of Candela (1992). This analysis needs to be extended to encompass the data collected since then. We also need to examine the application of basis functions that are more appropriate to the Bank topography than the polynomials currently in use. We need to be able to predict the tides reliably in order to use the data from shorter cruises where there are not enough data to perform a local tidal analysis.

• ii) For the broad-scale cruises and a few other cruises where there is sufficient spatial coverage of currents, we plan to make use of an adjoint version of the Dartmouth Circulation Model (in cooperation with D. Lynch and Chris Naimie) to provide detided, dynamically consistent, Bank-wide current fields with which to analyze shipboard observations in terms of flow pathways and advective time scales (Lynch, Naimie and Hannah, submitted). We also intend to use these interpolated/extrapolated current fields, in conjunction with the moored and drifter data, to examine the Bank-wide sub-tidal current regime and its seasonal variation.

• iii) There is a planned collaboration with Ashjian, Durbin, Wishner, and Irish to examine the spatial and seasonal plankton distributions where the ADCP backscatter data are able to provide reasonable descriptions of the plankton field. We will compare forward acoustic backscatter models with net tow results for the single-frequency ADCP data to document where the acoustic data can provide reliable results. We will also examine the several years of moored ADCP data from the long-term moorings on the south flank and Northeast Peak to look in detail at the temporal development of the plankton community. Lastly, a planned deployment of a NB-ADCP on the eastern portion of the north flank of the Bank will also be returning backscatter intensity data which will be studied for evidence of the hypothesized on-bank advection of calanus adults and copepodites from the Gulf of Maine during the late winter and early spring.

E. Contemporary and Retrospective Hydrographic Data Analysis (Flagg,leader)

We plan analysis of historical and contemporary hydrographic data to define the decadal scale climatology of the Bank and the Northeast Channel, thus providing an historical context for the GBS. Of interest are decadal changes in slope water characteristics and whether the analyses of Gatien (1976) and Wright (1977), among others, concerning the relative location and character of the Labrador and western, or warm, slope waters, still pertain. We will determine whether there have been any changes in the Bank, slope, flank, or channel hydrography from the historical norm. Boothbay Harbor data (courtesy of Mark Lazzari) show about a one degree temperature increase during the early phase of the GLOBEC program and a transition to some of the highest temperatures seen this century. Petrie and Drinkwater (1993) attribute much of the long-term temperature (and salinity) variability on the Scotian shelf and within the Gulf to changes in the Labrador Current transport and/or changes in entrainment with surrounding waters. Thus hydrographic changes over the slope may also govern conditions on Georges Bank as well.