Combining Satellite Ocean Color Imagery and Circulation Modeling to Forecast Bio-Optical Properties: Comparison of Models and Advection Schemes
Abstract
Remote sensing of ocean color provides synoptic surface ocean bio-optical properties but is limited to real-time or climatological applications. Many applications, including navy mission planning using electro-optical sensor performance models, would benefit from a forecast capability. To achieve this, we couple satellite imagery with numerical circulation models to provide short-term (24-48 hr) forecasts of bio-optical properties. These are first-order approaches; they do not account for any biogeochemical mechanistic processes (growth, grazing, sinking, resuspension), only dynamical processes (currents). Nonetheless, by comparing forecast distributions with next-day satellite imagery, we can assess errors and estimate how strongly the physical processes control the bio-optical distribution patterns. We compare optical forecast results from three Navy models and two advection approaches. The Intra-Americas Seas Nowcast/Forecast System (IASNFS), the Hybrid Coordinate Ocean Model (H YCOM), and the Northern Gulf of Mexico Nowcast/Forecast System (NGOMNFS) provide current direction and magnitude at hourly time-steps, at 6km, 4km, and 2km resolution, respectively. We apply the current vectors from these models to 1km resolution SeaWiFS-derived bio-optical properties (chlorophyll, backscattering coefficient, total and inorganic suspended particulate matter concentration) to produce advected, surface forecast images, using both a passive tracer advection scheme (Eulerian approach) and a particle trajectory/accumulation scheme (Lagrangian approach). Difference images between the next-day, satellite-derived optical fields and the model-advected fields provide a quantitative assessment of the forecast accuracy of the three models and two advection schemes, to assess the degree to which physical dynamics control the bio-optical distribution patterns. We compare different seasons (spring vs. fall) as well as different forecast periods (24 vs. 48hr).
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 2008
- Accession Number
- ADA503936
Entities
People
- Brandon J. Casey
- Dong S. Ko
- Peter M. Flynn
- Rebecca E. Green
- Regina D. Smith
- Richard W. Gould Jr.
- Robert A. Arnone
- Tamara L. Townsend
Organizations
- United States Naval Research Laboratory