HIGH RESOLUTION COASTAL CIRCULATION: MERGING MODELS AND OCEANCOLOR DATA*

摘要

Remotely-sensed ocean color provides a window into coupled biogeophysical dynamics at thesurface of the coastal ocean. Satellite imagery provides synoptic, real-time measurements over largespatial scales; however, typical spatial resolution (250m – 1km) may be limiting when considering somecoastal processes that evolve over much finer spatial scales (such as ~10m’s) . Futhermore, temporalresolution is generally lacking in remote sensing imagery with approximately one or two passes per day incomparison to coastal processes which occur on hourly (or less) scales. Thus, the underlying physicalprocesses that influence the observed optical distributions are often too complex and are manifest at spaceand time scales that preclude full understanding from imagery alone. Alternatively, advanced numericalmodels of the coastal ocean applied at spatial scales of 10 to 100 m incorporate relevant coastal dynamics(such as wind, tide, and river forcing, vertical mixing, advection and dispersion), complex shorelines andbathymetry, and are able to predict coastal circulation over very short time scales throughout the entirewater column. These high-resolution models are often exercised as a virtual laboratory for understandingthe cause and effect between forcing and circulation. When coupled, numerical models and remotelysensedobservations can provide a powerful tool to both understand and predict dynamical processes andoptical patterns in coastal waters. From a military perspective, remotely-sensed ocean color data is oftenthe only observational window on ocean processes in denied areas making the coupling of satelliteimagery with numerical models all the more imperative for developing a preditive capability in suchregions.Aside from a deepened understanding of coastal processes, merging numerically modeledcirculation with satellite imagery can result in improvements to the modeled circulation, particularly inintertidal regions where few observations are available to quantify the land-sea interface. Once a morecomplete understanding of how ocean color data and modeled circulation are coupled, an appropriate dataassimilation approach can be developed that will allow coastal ocean models to take full advantage of theobservational information presented in the satellite imagery.SeaWIFS and MODIS satellites provide daily products of the bio-optical and SST properties forcharacterizing and monitoring coastal conditions (Arnone and Parsons, 2004). Advances in inter-satellitecalibration and atmospheric correction have been used to provide quantitative analyses of the opticalproperties in coastal areas. These satellites can now be used to determine how optical characterisiticschange from image to image; they provide a quantitative means for defining changes which can resultfrom the physical forcing. Current algorithms used with these ocean color satellites are based on semianayticalapproaches which use spectral changes to uncouple the inherent optical properies of absorption(total, phytoplanton, detritus and Colored dissolved organic) and backscatteirng (Lee et al. 2002). Thebackscattering properties are strongly associated with the particle concentration (in addition to the size ,shape, index of refraction). Additionally, optical propeites such as the beam attenuation coefficient?