19 Forecasting the Coastal Optical Properties Using Satellite Ocean Color
347
and standard deviation) of the backscattering coefficient forecast were characterized
by the “forecast error”, and “persistence error”.
The forecast error showed higher errors in locations where there are strong currents and significant changes in the dynamics. We attributed these to uncertainty in
the physical processes of the model forecast. The statistics of the persistence error
(daily changes in satellite bio-optical properties), although appearing low, are not
representative of the changing bio-optical conditions since the number of satellite
observations used to generate the persistence statistics is not representative. We presented the forecast errors and persistence errors as absolute values. Perhaps a better
way to describe the error would be using a percent difference in the forecast error.
This may be more useful for managers using the forecast to look for maximum
changes.
Coastal managers and researchers require new methods similar to weather forecasting to assess the coastal ecological conditions on time scale of hours. We
demonstrated a new capability for using ocean color remote sensing to provide both
the initialization and validation for bio-optical forecasting. On short time scales the
physical processes are shown to be representative of the distributions of bio-optical
properties. Future capabilities in ecological forecasting will rely on some degree
of physical forcing in addition to bio-optical processes. We have shown an initial
capability for coastal optical forecasting. However, improved methods to characterize the forecast uncertainty which are used in weather forecasting can be applied to
bio-optical forecasting in the immediate future.
Acknowledgements This research was supported by the Naval Research Laboratory 6.1
“BIOSPACE ”program and NASA REASoN program “Integrated analyses of ocean products in
the Gulf of Mexico”. Thanks to reviewers for useful suggestions and comments.
References
Arnone RA, Casey B, Ko DS, Flynn P, Carollo L, Ladner S (2007) Forecasting coastal optical properties using Ocean Color and coastal circulation models. SPIE Optics and Photonics,
Conference 6680-43, San Diego, CA
Arnone RA, Flynn P, Ko DS, Martinolich P, Gould R, Haltrin V (2006) Forecasting optical properties from satellite derived optical properties. Proceedings of the Ocean Optics XVIII, Montreal,
Canada
Arnone RA, Parsons AR (2004) Real-time use of ocean color remote sensing for coastal monitoring. In: Miller RL, Del Castillo CE, McKee BA (eds.) Remote sensing of the coastal
environments, Chapter 14, Springer Publishing, Kluwer Academic, Dordrecht, pp. 317–337
Casey B, Arnone RA (2007) Simple and efficient technique for spatial and temporal composite
imagery. SPIE Optics and Photonics, Conference 6680-43, San Diego, CA
Jolliff JK, Kindle J, Shulman I, Penta B, Friedrichs M, Helber R, Arnone RA (2008) Summary
diagrams for coupled hydrodynamic – ecosystem model skill assessment. J Mar Sys
76(1–2):64–82
Ko DS, Preller RH, Martin PJ (2003) An Experimental Real time Intra Americas Sea Ocean
Nowcast/Forecast System for Coastal Prediction. AMS 5th Conference on Coastal Atmospheric
& Oceanic Prediction & Processes, New York
Ladner SD, Sandidge JC, Lyon PE, Arnone RA, Gould RW, Lee ZP, Martinolich PM (2007)
Development of finer spatial resolution optical properties from MODIS. Proceedings, SPIE
Optics and Photonics Meeting, San Diego, CA
347
and standard deviation) of the backscattering coefficient forecast were characterized
by the “forecast error”, and “persistence error”.
The forecast error showed higher errors in locations where there are strong currents and significant changes in the dynamics. We attributed these to uncertainty in
the physical processes of the model forecast. The statistics of the persistence error
(daily changes in satellite bio-optical properties), although appearing low, are not
representative of the changing bio-optical conditions since the number of satellite
observations used to generate the persistence statistics is not representative. We presented the forecast errors and persistence errors as absolute values. Perhaps a better
way to describe the error would be using a percent difference in the forecast error.
This may be more useful for managers using the forecast to look for maximum
changes.
Coastal managers and researchers require new methods similar to weather forecasting to assess the coastal ecological conditions on time scale of hours. We
demonstrated a new capability for using ocean color remote sensing to provide both
the initialization and validation for bio-optical forecasting. On short time scales the
physical processes are shown to be representative of the distributions of bio-optical
properties. Future capabilities in ecological forecasting will rely on some degree
of physical forcing in addition to bio-optical processes. We have shown an initial
capability for coastal optical forecasting. However, improved methods to characterize the forecast uncertainty which are used in weather forecasting can be applied to
bio-optical forecasting in the immediate future.
Acknowledgements This research was supported by the Naval Research Laboratory 6.1
“BIOSPACE ”program and NASA REASoN program “Integrated analyses of ocean products in
the Gulf of Mexico”. Thanks to reviewers for useful suggestions and comments.
References
Arnone RA, Casey B, Ko DS, Flynn P, Carollo L, Ladner S (2007) Forecasting coastal optical properties using Ocean Color and coastal circulation models. SPIE Optics and Photonics,
Conference 6680-43, San Diego, CA
Arnone RA, Flynn P, Ko DS, Martinolich P, Gould R, Haltrin V (2006) Forecasting optical properties from satellite derived optical properties. Proceedings of the Ocean Optics XVIII, Montreal,
Canada
Arnone RA, Parsons AR (2004) Real-time use of ocean color remote sensing for coastal monitoring. In: Miller RL, Del Castillo CE, McKee BA (eds.) Remote sensing of the coastal
environments, Chapter 14, Springer Publishing, Kluwer Academic, Dordrecht, pp. 317–337
Casey B, Arnone RA (2007) Simple and efficient technique for spatial and temporal composite
imagery. SPIE Optics and Photonics, Conference 6680-43, San Diego, CA
Jolliff JK, Kindle J, Shulman I, Penta B, Friedrichs M, Helber R, Arnone RA (2008) Summary
diagrams for coupled hydrodynamic – ecosystem model skill assessment. J Mar Sys
76(1–2):64–82
Ko DS, Preller RH, Martin PJ (2003) An Experimental Real time Intra Americas Sea Ocean
Nowcast/Forecast System for Coastal Prediction. AMS 5th Conference on Coastal Atmospheric
& Oceanic Prediction & Processes, New York
Ladner SD, Sandidge JC, Lyon PE, Arnone RA, Gould RW, Lee ZP, Martinolich PM (2007)
Development of finer spatial resolution optical properties from MODIS. Proceedings, SPIE
Optics and Photonics Meeting, San Diego, CA