Attendees: Francisco Chavez, Robert Maffione, Yi Chao, Yuhe Song, Jim McWilliams, Fei Chai, Gernot Friederich, Ken Johnson, John Kindle, Baldo Marinovic, John Ryan, Jeff Paduan, Andrew De Vogelaere, Brian Schlining, Mario Tamburri, Reiko Michisaki, Mike Matthews (email addresses)

1. Discussion: Mission / Links with the MBNMS: Andrew De Vogelaere, Mario Tamburri
2. Discussion: Physical Modeling: Jim McWilliams, Jim Kindle, Yi Chao
3. Discussion: Biological Modeling: Francisco Chavez, Fei Chai, Yuhe Song
4. Discussion: Bio-optical Modeling: Robert Maffione
5. Discussion: For coupled modeling, what temporal scale do we begin with (seasonal, interannual)? Which modeling questions do we begin with?
6. Discussion: Observations: Francisco Chavez, Reiko Michisaki, Jeff Paduan, Gernot Friederich, Ken Johnson, Baldo Marinovic
7. Discussion: Data Assimilation for Modeling
8. MUSE II

Andrew DeVogelaere summarized the goals and roles of the MBNMS management and how he expects this NOPP-funded research will benefit the Sanctuary. The expected benefits include:

• Compilation of information
• Response to requests for information
• (Understanding) transport of introduced species
• Locating (defining) protected areas
• Ecological studies
• Prediction

It was agreed that these expectations are within the scope and time frame of the proposed research . Mario Tamburri provided a summary of a recent MBNMS workshop to develop priorities and approaches.

John Kindle summarized specifics of the nested modeling at NRL, including resolutions and data assimilation. Jim McWilliams and Yi Chao summarized ROMS nested modeling. The current resolution of the regional model of the eastern North Pacific in ROMS is 5 km; the next step for this effort will be ~ 1km. Vertical resolution is ~ 5 m, sufficient to resolve crucial processes and variation in these waters. Synoptic realizations of model results show remarkable consistency with circulation features observed in situ. Jim McWilliams emphasized that simply due to the chaotic nature of circulation, we cannot expect to compare specific synoptic realizations of model and observations. That is, although the model may correctly represent the characteristic processes and scales of the circulation, we cannot expect to match model and observation exactly in space and time. Comparisons should focus on statistics of model and observed fields, e.g. EOFs of time series.

Fei Chai summarized the 10-component ecosystem model developed for the equatorial Pacific and the expected modifications for the study region, including the addition of iron and oxygen. This model has been coupled to a physical circulation model of the northeast Pacific (the region around Van Couver Island). Coupling to ROMS is underway with Fei and Yi.

Robert Maffione summarized fundamentals of radiative transfer relevant to the coupled modeling and a specific approach for modeling the underwater light field for physical and biological components. Robert considers spectral resolution, in addition to depth resolution, to be very important and not too computationally expensive. It was agreed that Robert and Francisco would provide the software for spectrally resolved computations.

For all modeling efforts, accurate insolation is essential. Francisco suggested use of the recently developed PAR product produced by the SeaWiFS project. This would include the most important factor influencing incident light: cloud cover. John Ryan will obtain and make available these PAR fields.

1. Wind (Quick scatterometer)
2. Tide (Spatial and time)
3. Current
4. Biogeochemical (any)
5. Temperature and Salinity
6. Sealevel (T/P)
7. Surface currents (CODAR)
8. PAR

• Iron
• Light

• Bacteria, DOC

• Sediments
• Process below euphotic zone
• Zooplankton
• Carbon

• Oxygen
• Currency (carbon/chl ratio; N/chl ratios)

• 2001: Develop an adjoint model for ROMS
• 2002: Testing ROMS and its adjoint for the Monterey Bay configuration
• 2003: Estimate parameters in the coupled physical-biogeochemical model

• Monthly, if possible, via conference call.
• Once every six months with a subset of observational and modelers
• Once annually all partners, 2 day meeting
• Alternate meeting sites between LA and Monterey

• John Ryan: Investigate SeaWiFS PAR
• Robert Maffione & Francisco Chavez: Develop code for modeling E & E₀
• All: re: NOPP web page (http://www.mbari.org/bog/NOPP), send URLs to reiko@mbari.org
• MBARI: Summarize expectations of all groups

The primary focus of the group will be around the ROMS physical model and the ten component ecosystem model. A parallel effort will be carried on by John Kindle and his group at NRL and what is learned as a result of the ROMS focus will be incorporated into the Navy models. There are several ongoing ecosystem model developments (UCLA, NCAR, NRL etc.) that will be considered during the course of the grant. Given the first sentence these are the near term tasks for each of the groups:

1. The physical modelers (McWilliams, Chao) will develop a three component (basin, regional, local) one-way nested model. They will develop means to analyze lagrangian trajectories from model simulations. Questions and comments: What are the resolutions of each domain? Where will the simulations for this project be run (UCLA or JPL)? Is it possible to keep the UCLA and JPL versions concurrent? During the October 5 conference call we decided that tentative answers to these questions and task assignments for the physical modeling would be circulated prior to the November 2 conference call. Based on this and other input the management office would create a timeline with milestones. This would provide a means for gauging progress.
2. The assembly of atmospheric forcing fields for the physical models back to 1990 will be the responsibility of John Kindle and Yi Chao. Chao and Ryan will work on QuickScat winds. Should there be a study of model output using pre-COAMPS, COAMPS and Quick Scat winds? From October 5 conference call: The pre-COAMPS fields include the NCEP re-analysis, Fleet numerical products and in the future ECMWF re-analysis. These will drive the basin scale (always) and regional and local models (pre 1998 for COAMPS, pre 1999 for QuickScat). Uncertainties exist about buoyancy fluxes and relative humidity field data will be investigated. The desire to recruit a meteorologist into the group was expressed. Kindle will provide a short summary prior to November 2 conference call. MBARI will explore the best insolation/irradiance product for driving heating and primary production in the model. Access to model (i.e. COAMPS) data will be required. Also investigating NASA PAR products. Comparisons between products and observations (moorings) will be made. Ryan to report on progress on November 2.
3. The ecosystem modelers (Chai, Barber, Chavez) will integrate the 10-component ecosystem model into ROAMS. The first effort was to integrate a 9-component model and CO2 will be added next since the code is there and there is a large validation data set. Iron (Chavez, McWilliams) and oxygen (Chai) will be added next. Bacteria and DOC will be considered at a later date. Questions and comments: Will the ecosystem model be in all of the model domains? From October 5 conference call: Eventually the ecosystem model will run in all domains. The first exercise will be to incorporate the 10 component model into the regional domain to compare with UCLA simulations using a simpler ecosystem. MBARI will provide an alkalinity salinity relationship for the region for CO2 calculations. Further complexity will be added to the carbon chemistry at a later stage. Chai to provide update prior to or on November 2.
4. A subroutine will be developed by Maffione and Chavez that will take the output from the ecosystem model and estimate net and scalar spectral irradiance, and inherent optical properties in the water column. Net spectral irradiance will be used to compute solar heating of the water column, and spectral scalar irradiance will be used to compute primary productivity rates. Maffione to provide update on November 2..
5. Chao and Sung will develop data assimilation methods for physical (TOPEX, SST, mooring T, S , u, v) and biological (ocean color, mooring chlorophyll, backscatter, Ed, Lu) properties. Please give estimates as to when we will be using data assimilation. Candidate to perform work identified and will hopefully begin shortly. No tangible results expected in the first year.
6. Validation of the simulations will require a joint effort. We will need to decide (soon) what fields and statistics will be produced from the models and observations. Paduan, McWilliams and Chai will lead validation of the physical model, Chavez, Chai, Marinovic and Barber will lead validation of the ecosystem model and Maffione and Chavez will lead validation of the bio-optical model. Our first goal will be to produce a simulation of the seasonal cycle for the Monterey Bay region. We will focus on the decade of the 90’s given the available observations. After the seasonal cycle we will work in two directions: a) smaller scales focusing on simulation of the period of the MUSE experiment (August, 2000) and b) interannual variability focusing on simulation of the 1997-98 El Niño and the 1999 La Niña. MBARI will begin assembling monthly/seasonal statistics for the region. Will assemble consistent fields for MB and CalCOFI. Ryan will report on progress on November 2.

top

return to Main page

Notes compiled by John Ryan, Reiko Michisaki (10/10/00)