JGOFS SMP PI, May 20-22nd, 1998 NCAR Boulder CO
The inaugural U.S. JGOFS Synthesis and Modeling Project Principal Investigator meeting was held at NCAR in mid-May, and a synopsis of the meeting is presented below.
Upcoming U.S. SMP events include:
July 1998: SMP Summer Workshop, Marine biogeochemical response to climate change, Durham NH
August 1998: Deadline for second round of SMP proposals
Spring 1999: Annual SMP PI meeting, Location TBD
July 1999: SMP Summer Workshop, Location and topic TBD
The first part of the meeting consisted of a series of brief overview talks for each of the funded SMP projects. PIs were requested to highlight the scientific goals of their project, the types of data sets and models involved in the analysis, the expected products (e.g., synthetic data sets; model parameterizations; model solutions), and the connections to the overall SMP program. The PI driven projects were broadly grouped into four topics: mass balance, extrapolation, and data analysis (Hansell, C. Keeling, R. Keeling, Sabine); process models and synthesis; regional models and studies; and basin and global models. Individual SMP project summaries are also available on the JGOFS web page.
SMP Working Groups
The most important outcome of the meeting was the formation of SMP working groups. When the topic of SMP organization was discussed, it was generally agreed that some common activities or working groups would facilitate the success of the SMP and that working groups will form and disband as needed. Two broad types of working groups (infrastructure and science) were proposed. The infrastructure groups will focus on constructing a common structure for sharing observational and synthetic data, completing diagnostic calculations, and comparing and validating models with data. They include:
-global-scale biogeochemistry models (Ed Laws and Ray Najjar)
-regional ecosystem model testbeds (Rob Armstrong)
-coastal--open ocean interactions (Frank Muller-Karger)
-large-scale data sets (Ralph Keeling and Chris Sabine)
-community synthesis and modeling (Mick Follows and Scott Doney)
The science working groups are more focused on specific issues and could form, for example, to write a comprehensive review paper. Two such groups were proposed:
-nitrogen fixation (Doug Capone)
-satellite biogeochemistry (Mary-Elena Carr).
The working groups met briefly at the meeting and were tasked with writing specific terms of reference for the next 6 to 12 months, which will then be posted on the SMP web site for discussion.
The SMP Science and Implementation Plan (Sarmiento and Armstrong, 1996) defined a broad, ambitious program, and a number of elements are not well covered by the present SMP group. These include POM and DOM processes (e.g., mechanisms of DOM production, grazing and sloppy feeding, variable Redfield ratios, subsurface particle remineralization, microbial/DOM interactions); biogeochemical functional processes (e.g., denitrification, calcification and water column dissolution, silica cycling); trace metal/biological interactions (e.g., Fe chemistry and modeling, aeolian input), and sediments. Strong concerns were also raised that with the decline in DOE and NOAA funding for the marine carbon system that the synthesis of the global survey CO2 (particularly the Pacific) could be left incomplete unless funded under the SMP. Several other topics related to but not directly part of SMP were raised but the consensus was that these issues were best dealt with by inviting experts in these areas either to PI meetings or summer workshops as needed. Theses topics included: margins/riverine input; air-sea gas exchange; climate and inter-annual variability; southern ocean and Arctic processes.
There was also some discussion about the need to define a common language or set of terms. For example, how should we define "net community production".
Draft Data-Model Policy
The open data policy JGOFS field programs where PIs must submit their data to a central data management office and make the data available on the web after some embargo period has been quite successful, and there was general agreement that the SMP should develop a similar, reciprocal "model-data" policy appropriate for synthetic data and model results. The broad goals of the policy should acknowledge the rights of individual researchers to their work while allowing for scientific replication of results, access of the scientific community to synthetic data and model simulations, and development of community models and code/data resources. Elements of a minimum model-data policy would include thorough documentation of equations and numerical algorithms in the literature or technical reports and access, after an appropriate embargo period, of synthetic data sets and model forcing, boundary conditions, validation data and final model solutions. PIs would also be encouraged to share their model codes. Depending on the quantity of model-data, web-based access would be supported either from a central archive or at the location of the PI. The SMP co-coordinators will draft a model-data policy based on existing JGOFS, GLOBEC and COoP polices and circulate it to the JGOFS Steering Committee for comments and approval.
The success of the SMP hinges on the availability of appropriate JGOFS and historical data sets for synthesis and model validation. Concerns were raised about the timing of the release of the AESOPS data, in particular, and coordinators offered to discuss this matter with the coordinators of AESOPS. The role of the Woods Hole data management office as a repository for models, synthetic data sets and historical observations was also discussed. Several people noted the existence of significant amount of relevant data from the International JGOFS and other programs (e.g., WOCE).
Connections to Other Programs
Dave Siegel briefed the group on a recent meeting of the International JGOFS North Atlantic Synthesis team led by Mike Fasham. The North Atlantic group set out a three level synthesis plan: a special issue of Deep-Sea Research II of invited review papers; development and distribution of synthetic data sets; and international collaboration on specific synthesis topics (e.g., CO2 system; NABE/Biotrans; deep-sea traps; biogeographical provinces).
Regarding other programs (WOCE, EOS, SeaWiFS, Carbon Modeling Consortium), it was noted that there is substantial overlap at the PI level, and it was suggested that informal contacts could be maintained by individuals within SMP.
Overall SMP Synthesis
The topic of the overall synthesis of JGOFS was introduced, and the following stimulating discussion showed that there is a wide range of concepts for what form this should take, ranging from a book of review papers to a "how-to" biogeochemistry modeling cookbook to an actual set of community numerical models. A specific working group was formed to further develop these ideas and bring them back to the SMP PI group for more discussion.