2.1 Review of Status of Science Plan

The two major JGOFS goals as embodied in the International Science Plan (JGOFS Report No. 5, SCOR, 1990) are:

1. To determine and understand on a global scale the processes controlling the time-varying fluxes of carbon and associated biogenic elements in the ocean, and to evaluate the related exchanges with the atmosphere, sea floor, and continental boundaries.

2. To develop a capability to predict on a global scale the response of oceanic biogeochemical processes to anthropogenic perturbations, in particular those related to climate change.

The sum of these goals lays out as the central aim of the JGOFS program the embodiment of an improved understanding of ocean fluxes of carbon and associated biogenic elements in the form of models. The discussion of modeling in the U. S. JGOFS Long Range Plan (U. S. JGOFS Steering Committee, 1990) begins with a brief summary of the distribution of biologically active chemical compounds in the ocean, concluding that "these observations challenge us to develop models which couple the physical forcing, biological agents of transformation, and chemical substrates so that theories about the ocean's production system can be formulated and tested. Through large-scale models the theories found to work in areas of study can be applied far afield in regions remote to study sites. The legacy of U. S. JGOFS will be the understanding of the system, as encoded in algorithms and models, which will enable us to monitor the state of the ocean in real-time and to predict its future course in an era of climate change." This section of the implementation plan focusses on the combined data synthesis and modeling activities that will be required to meet the overall goals of JGOFS as summarized in these two documents.

The large temporal and spatial dimensions of these goals require dedicated effort to draw together the collective understanding that the JGOFS project is designed to attain. Indeed the argument used to justify incremental funds to pursue components of the U.S. Global Change Research Program, like JGOFS, is that this collective understanding will be greater than the sum of the individual elements. It is unrealistic to assume that all scientists engaged in field aspects of these projects will have comparable interest in or shared commitment to this responsibility. For some participants in JGOFS the individual investigator's primary results of experimentation and observations that make up the process studies, time series stations, and global surveys are of foremost interest. For others the main intellectual challenge is in understanding a larger scale phenomenon, like the central focus of a process study, i.e. a seasonal or interannual event. At the interface of biology and geochemistry within ocean science, there has been relatively little opportunity to develop syntheses and models that approach the scale of the JGOFS effort. No prior project has rivaled the interdisciplinary scope, global coverage, and U.S. investment in JGOFS. Thus the comprehensive scope of the JGOFS science plan requires an unprecedented and systematic commitment to synthesis and modeling.

All investigators play a critical role in the ensemble of synthesis activities when they fully analyze their individually funded efforts and submit these data to the JGOFS data base. In addition, many process study grants were funded with the specific expectation that groups of investigators would be working together to interpret multiple linked data sets. However, the more complete synthesis and modeling called for in JGOFS has not been specifically planned and it is too important to be left to serendipity.

Modeling activities are critical to the success of JGOFS synthesis. Model representations are the means by which we will test the products of integration among process studies, time series stations, and global surveys. Modeling will be required to fully assimilate the understanding gained through JGOFS in comprehensive analyses of the ocean's role in the global carbon cycle.

The current Mid-Program Strategy augments the effort outlined in the science plan, making specific recommendations for the latter half of the U.S. JGOFS program regarding the role of data synthesis and modeling in current and future field programs, the priority topics on which efforts should be directed, the hierarchy of data synthesis and modeling work required, and the balance of funding between data synthesis and modeling and observations. Investment in comprehensive analyses of the process and time series data sets is a necessary precursor to the larger scale integration of JGOFS findings. The suite of model subcomponents and cadre of scientists who will ensure fulfillment of the JGOFS promise must be developed during the next few years.

2.2 Review of Current Activities

A relevant perspective on the status of synthesis activities is the experience with studies that have been carried out up to the present. It is very clear that too little synthesis and modeling activity was supported for either the NABE pilot project or EqPac. However, because of these experiences we have a clearer view as to how to proceed with the remaining large process studies. Workshops and grants of longer duration are essential, especially when the field campaign schedule is as demanding as it has been for previous processs studies.

As data records for both HOT and BATS approach half a decade, it is also appropriate to review mechanisms in place and assess need for support of synthesis and modeling efforts with these data. Some ancillary science projects have been funded in association with the time series stations, and greater efforts are being taken to encourage other investigators to use JGOFS time series data. It has become very clear from reviews of these data that in many instances where the North Pacific and Bermuda gyres were thought to have little seasonal to interannual variability there is now evidence to the contrary.

As the two global survey components of JGOFS, the CO2 survey and SeaWiFS mission, reach maturity, they too will have requirements for synthesis and modeling activities. As with the process studies, some of these will be self contained, but others will certainly link to the results of process studies and time series stations.

The modeling effort supported directly through the U. S. JGOFS program has been very limited. However, there has been a considerable amount of modeling research relevant to U. S. JGOFS goals funded outside the direct U. S. JGOFS umbrella. The following summary of current U. S. activities is broken down into four major categories beginning with a discussion of modeling research in support of process studies, followed by model studies in support of the time series stations and the global survey, and concluding with a discussion of model development efforts independent of the field programs.

Process Studies

The North Atlantic Bloom Experiment (NABE) was not preceeded by any basin scale model studies. However an eddy resolving diagnostic model was used to provide useful guidance at the time of the expedition through real time simulations of the eddy field in the North Atlantic. This same model is being used to carry out biological simulations. Additional studies have been carried out with a non-eddy resolving model making use of the Fasham, et al. [1990] ecosystem model in a seasonal GCM of the North Atlantic. Results from this coupled model have been compared to satellite data (Sarmiento, et al. [1993]) as well as time series observations at Bermuda Station S and Ocean Weathership Station India (Fasham, et al. [1993]). Plans exist to make use of an updated version of this model for analysis of NABE observations. Additional work is being undertaken to emplace an ecosystem model into the WOCE eddy resolving community model of the North Atlantic. The ongoing studies will play in important role in planning the next North Atlantic study to be done later this decade.

A GCM ecosystem model of the Equatorial Pacific played a major role in planning for the Equatorial Pacific Study (EqPac). Several modeling studies are currently underway to study the physical and meteorological regulation of oceanic primary production in the Equatorial Pacific. The Arabian Sea Study has also had an important circulation model component that was used in planning and will be utilized in interpretation of results. The Southern Ocean process study was recently initiated with a request for modeling proposals by the NSF.

It is important to note that none of the models discussed abeve succeeds in capturing all the important features of the observations. There is much work that remains to be done.

Time Series Stations

A number of groups have been working with the data being obtained from the Bermuda and Hawaii time series stations (BATS and HOTS, respectively). Some studies have focussed on the development of more realistic physical models for these regions, wheras others have focussed on developing improved ecosystem models.

High temporal physical and bio-optical resolution time series have been collected from other sites in the Sargasso Sea ( Dickey et al., [1993] ) as well as the North Atlantic south of Iceland ( Dickey et al., [1994] ; Stramska and Dickey [1994] ). These data sets are being used for models at present (e.g., Stramska and Dickey [1994] ) and should be useful to the JGOFS modeling community at-large.

A number of workshops have been organized to bring together scientists who are interested in various aspects of biological/physical modeling at time-series stations. The enthusiastic response to these meetings reflects the broad community interest in time-series studies.

Global Survey

The global survey of carbon is being carried out on WOCE ships primarily with DOE support, as well as on NOAA ships with NOAA support. Numerous scientists within that study have been working on the interpretation of these observations using inverse modeling methods aimed primarily at determining the meridional transport of carbon by the ocean circulation.

Model Development

The development of models requires an ability to predict the circulation, as well as the formation and remineralization of organic matter. A great deal of synthesis and modeling research has been carried out on the formation, transport, and remineralization of organic matter in association with the process study and time-series stations. Although many papers have been written reporting the observations obtained in the process studies, there is still a major need for synthesis papers that provide an overview of the important results. Such syntheses are essential if the observations are to be utilized by modelers.

2.3 Mid-Program Implementation Strategy

U. S. JGOFS must begin a substantial ramp-up in synthesis activities and the development of models. This is necessary to provide adequate synthesis and modeling support for ongoing and future field programs and time series stations, as well as to meet the goals for grand syntheses and models envisioned at the inception of the program. Doing this will require that the goals of the synthesis and modeling program be defined more specifically. A number of suggestions for this are made below, but more are needed. The Steering Commitee must continue to evaluate progress and identify needs. Resources need to made available to hold regular meetings for synthesizers and modelers and key observationalists to let each other know what they are doing and to make plans for future work. Also, there should be a summer program including students and senior investigators. An early goal of these programs would be to further define the synthesis and modeling goals given below, and to begin addressing specific topics such as the modeling of time series stations, studies of the process study sites, and specific processes of importance such as the cycling of POM and DOM. Such a program would give investigators a chance to interact with each other and to train and motivate a future much needed cadre of scientists.

Following is a first suggestion at defining more specific synthesis and modeling goals within the context of the JGOFS observational program that would improve our understanding and ability to model each of the components of an ocean flux model. The list is given in approximate priority, with ongoing field programs receiving higher priority for early support, and longer term/larger scale synthesis and modeling receiving priority for later support.

1. Each process field program should have a synthesis and modeling activity directly associated with it. The only remaining process study planned for the field study phase of JGOFS is the Southern Ocean Study. However, future field studies will develop out of the synthesis and modeling phase of the program, such as the North Atlantic carbon budget study and Bermuda Time Series Station control volume experiment that have recently been proposed. In order to support these field programs we need:
   • Inverse models and synthesis efforts based on process study observations. A particular need is for syntheses that highlight the important results of these complex studies.
   • Simpler models focussed on processes that are important in a given region.
   • Ecosystem and chemical cycling models based on basin scale ocean circulation models.
   • Eddy resolving models for both interior processes and to represent basin-margin interactions.

2. Time series stations should have a synthesis and modeling activity directly associated with them. We need
   • Inverse models and synthesis efforts based on time series observations. A particular need is for syntheses that highlight the important results of these studies.
   • Simple mixed layer and one-dimensional models that make it possible to explore a wide range of chemical and biological sub-models.
   • Mesoscale models capable of taking horizontal processes into consideration. Both BATS and HOTS are showing evidence of significant horizontal inputs that need to be addressed.

3. Synthesis and modeling of global survey observations from WOCE hydrographic sections is required. We need
   • estimates of meridional nutrient and CO2 fluxes.

   • estimates of the spatial distribution of air-sea CO2 exchange.

   • synthesis studies highlighting the important results.

4. The specific components of an ocean flux model need to be addressed. These topics need not necessarily be addressed within the context of a given field program. The basic elements that make up an ocean flux model are:
   • Models of ocean circulation with a temporal and spatial resolution that is appropriate to the problem under consideration. Such models include 1-D models for use at time-series stations, global and basin scale GCM's, nested high-resolution models of ocean margins within basin-scale GCM's, and eddy resolving models.

   • Euphotic zone production. Models which describe the flows of carbon, nitrogen, and energy between elements of a food web.

   • The transport of DOM and POM out of the regions where they are produced to the places where they are remineralized.

   • A model for the remineralization of DOM and POM. A more specific listing of processes that must be studied includes:
     • processes that control the surface nutrient and total carbon concentration such as in the High Nutrient Low Chlorophyll (HNLC) regions of the North and Equatorial Pacific and Southern Ocean
     • improved understanding of the role of micronutrients such as iron.
     • processes that control seasonal behavior
     • processes that control the composition of the material formed at the surface, e.g., the Redfield ratio, the ratio of calcareous versus silicious organisms, and the proportion of dissolved versus particulate organic matter (DOM vs. POM, respectively). A major strategy for the development of carbon cycle models is to base them on the cycling of nutrients such as phosphate and nitrate. However this requires knowing what the stoichiometric ratios of uptake are.
     • relationship between primary and new production and processes controlling the rates of these.
     • stoichiometric ratio of remineralization of DOM and POM at various depths.

5. Syntheses of satellite data and models for the interpretation of satellite observations, such as efforts which utilize remotely-sensed ocean color data to predict primary and new production through the use of algorithms, and/or through assimilation into ecosystem/GCM models.

6. Simulations are required that address how ocean fluxes will respond to the anthropogenic CO2 transient, and the climate and ocean circulation changes that will occur in response to greenhouse warming. These would make use of the models developed under (4), but would include in addition the use of coupled ocean-atmosphere models for predicting future ocean circulation changes. An important constraint on such models is an ability to predict past changes in chemistry, thus the use of JGOFS findings to better interpret paleoceanographic and paleoclimate records should be supported.

The resources that will be required to develop the foregoing program are substantial. Much of these will be provided by funding agencies other than NSF, with a considerable stake in these problem areas by DOE, NASA, and ONR. A considerable amount can also be achieved by close collaboration with the GLOBEC modeling effort. However, it is essential that the NSF funded portion of the U. S. JGOFS program ramp up substantially in order to meet the goals of JGOFS.

We need to have an additional 5 to 10 scientists entrained into modeling and synthesis activities now, and we estimate that by the year 2000 the synthesis and modeling activity necessary to meet the JGOFS goals will require support for at least 10 to 15 scientists, with a comparable effort involved in associated field programs. Directed postdoctoral programs such as the present UCAR program will be of use in providing an influx of newly trained scientists, but it is clear from the response to previous synthesis and modeling initiatives within JGOFS that the capacity to initiate this task on a large scale is already in place today.

We propose specifically that the present NSF modeling support of $260k/yr should be incremented by $500k by FY96 for support of synthesis and modeling. The support of 10 to 15 groups will require a continued ramping up to the level of $3000k-4000k by the year 2000, with most of the increase occuring after the field programs are completed. A comparable level of support will be required for associated field work.

REFERENCES

Dickey, T., T. Granata, J. Marra, C. Langdon, J. Wiggert and etc., Seasonal Variability of Bio-optical and Physical Properties in the Sargasso Sea, Journal of Geophysical Research, 98, 865-898, 1993.

Dickey, T., J. Marra, M. Stramska, C. Langdon, T. Granata, R. Weller, A. Plueddemann and J. Yoder, Bio-optical and physical variability in the sub-arctic North Atlantic Ocean during the spring of 1989, J. Geophys. Res., 99, 22,541-22,556, 1994.

Fasham, M. J. R., H. W. Ducklow and S. M. McKelvie, A nitrogen-based model of plankton dynamics in the oceanic mixed layer, Journal of Marine Research, 48, 591-639, 1990.

Fasham, M. J. R., J. L. Sarmiento, R. D. Slater, H. W. Ducklow and R. Williams, Ecosystem behavior at Bermuda Station "S" and OWS "India:" a GCM model and observational analysis, Global Biogeochemical Cycles, 7, 379-416, 1993.

Sarmiento, J. L., R. D. Slater, M. J. R. Fasham, H. W. Ducklow, J. R. Toggweiler and G. T. Evans, A seasonal three-dimensional ecosystem model of nitrogen cycling in the North Atlantic euphotic zone., Global Biogeochemical Cycles, 1993, 417-450, 1993.

Stramska, M. and T. D. Dickey, Modeling phytoplankton dynamics in the northeast Atlantic during the initiation of the spring bloom, Journal of Geophysical Research Letters, 99, 10,241-10,254, 1994.