Multidisciplinary University Research Inititaive (MURI)
Principal Investigator for Rutgers: Oscar Schofield
AWARD N00014-06-1-0739
TOTAL AWARD $3,081,829
Project Dates: August 1, 2006 - April 30, 2009
The MURI project is funded by the Office of Naval Research
DOWNLOAD Full Proposal

Rutgers is collaborating with the following institutions:

• Dalhousie University - Katja Fennel
• Callifornia Polytechnic University - Mark Moline
• Woods-Hole Oceanographic Institute - Dennis McGillicuddy, Glenn Gawarkiewicz
• North Carolina State University - Ryoing He

Real-time data streams and environmental forecasts are used to optimize Naval operations by providing insight to how the environment will impact fleet operations.  In the open ocean, broad oceanographic features typical of the deep sea can be reasonably described by models assimilating satellite estimates of sea surface temperature (SST) and sea surface heights (SSH) combined with any available in situ observations that provide vertical information; however this is not the case for the littoral ocean where the dynamics are complex and prominent hydrographic features are often not resolved using SST and SSH.  Until new sustained spatial data streams are available and the data assimilative models capable of using them are developed our ability to nowcast and forecast ocean conditions on continental shelves will remain an elusive goal.  This is unfortunate as many areas of highest naval interest are broad western boundary continental shelves and shallow semi-enclosed seas. 

On continental shelves the optical properties are complex representing variable contributions of phytoplankton, Colored Dissolved Organic Matter (CDOM), and non-algal particles.  These constituents reflect the water column biology and chemistry, driven by the physical forcing, and are very effective at detecting hydrographic features not evident in SST or SSH data (Figure 1).  These optical properties are particularly effective at distinguishing minor density discontinuities where biological communities tend to accumulate. Numerous biological models of varying complexity are available to describe the dynamics that underlie the variability in ocean color (Ishizaka 1990, Lawson et al. 1995, Fasham and Evans 1995, Matear 1995, Robinson et al. 1996, Bissett et al. 1998, 2005); however the issue of how complex the biological models need to be is still unresolved.  This is especially true for continental shelves where a significant fraction of the organic matter is remineralized and recycled by the foodweb, with concurrent feedbacks on the optical properties.  In the recent past, limited availability of data and biological data assimilation approaches have precluded predictive skill experiments of the coupled physical-optical models.  Given the recent advances in ocean observation capabilities it is now possible to collect spatially extensive ocean color data appropriate for data assimilation (Glenn and Schofield 2003).  To this end, we propose to develop a data assimilative physical-optical modeling-observation system consisting of an ensemble of optical models of varying complexity in order 1) to improve our predictive skill for forecasting ocean color and 2) improve physical models by using ocean color to discriminate hydrographic features not detected using traditional data streams.  We will study the regulation of ocean color for a broad western boundary continental shelf with a specific focus on regions of high optical variability (fronts), which coincide with regions of high acoustic uncertainty.

• Develop and refine our ability to assimilate both in situ and remotely sensed physical and optical data into a ROMS (Regional Ocean Modeling System) model of the Mid-Atlantic Bight, using 4-D variational assimilation, based on tangent linear and adjoint versions of the physical model, and a suite of coupled bio-optical models tailored to this project.  Variational assimilation of optical data from the network will allow better estimation of the optical model state variables for initialization of medium-range forecasts (3-4 days).

• Develop an observational scheme to fuel the data assimilative model by accessing the international constellation of satellites (thermal imagers, ocean color sensors, and synthetic aperture radar), a multistatic nested CODAR array, and an ONR fleet of coastal AUVs (Gliders and REMUSs) outfitted to measure the in situ physics and optics.  Development includes the processing algorithms to invert bulk optical property into constituent components for assimilation and validation of the optical models.

• Use the coupled model to study the synergy between the ocean physics and optics on the Middle Atlantic Bight (MAB).  Ensemble bio-optical models will access our predictive skill in forecasting ocean color.  Efforts will focus on optically characterizing the sources and sinks of colored material and their rates of change on the shelf and using the fields to objectively define hydro-optical fronts.  Minimizing the model errors and uncertainties will be achieved through adaptive sampling with swarming AUVs.