estimatingthediffuseattenuationcoefficientfrom landsat（estimatingthediffuseattenuationcoefficientfrom landsat）.docVIP

Estimating diffuse attenuation coefficient from Landsat imagery in the Florida Keys: Implications for water quality assessment of coral reef ecosystems Palandro, D., Hu, C., Andréfou?t, S. and Muller-Karger, F. E. College of Marine Science, University of South Florida 140 7th Avenue South, Saint Petersburg, FL 33701, USA Tel: 727-553-1186, Fax: 727-553-1103 E-mail: palandro@seas.marine.usf.edu Submitted as a Research Note to the Special ICCB7 Proceedings Issue of: Hydrobiologia Keywords: remote sensing, satellite, water clarity, classification This paper has not been submitted elsewhere in identical or similar form, nor will it be during the first three months after its submission to Hydrobiologia. ABSTRACT The diffuse attenuation coefficient, K (m-1), describes the effective attenuation of light in the water column, therefore representing water clarity. The water column correction of remotely sensed imagery requires K in the mapping of shallow benthic habitats. Traditionally, K is derived from imagery using known locations of the same bottom type located at different known depths, assuming that light attenuates exponentially with depth. This technique is applied to three Landsat 7 satellite images (February, May and July 2000) for the Florida Keys Reef Tract between Key Largo to Key West. Interpolated depth, initially from NOAA vector data, with uncertainties of ±0.5m, was draped over the 30m spatial resolution Landsat data. K was derived for 27 sites where bright sand could be observed at varying depths between 3m and 20m. Results show that the blue (band 1) and green (band 2) bands provided the best information and were relatively consistent with time and location. The red (band 3) band provided more ambiguous results with several negative K values, attributed primarily to two combined factors: use of a simple atmospheric correction and the rapid attenuation of the red band in water. Despite the fact that these observations are a snapsh