No abstract

Theory and seaborne measurements are presented for the near infrared (NIR: 700-900 nm) water-leaving reflectance in turbid waters. According to theory, the shape of the NIR spectrum is determined largely by pure water absorption and is thus almost invariant. A ''similarity'' NIR reflectance spectrum is defined by normalization at 780 nm. This spectrum is calculated from seaborne reflectance measurements and is compared with that derived from laboratory water absorption measurements. Factors influencing the shape of the similarity spectrum are analyzed theoretically and by radiative transfer simulations. These simulations show that the similarity spectrum is valid for waters ranging from moderately turbid (e.g., water-leaving reflectance at 780 nm of order 10 Ϫ4 or total suspended matter concentration of order 0.3 g m Ϫ3 ) to extremely turbid (e.g., reflectance at 780 nm of order 10 Ϫ1or total suspended matter of order 200 g m Ϫ3). Measurement uncertainties are analyzed, and the air-sea interface correction is shown to be critical for low reflectances. Applications of the NIR similarity spectrum to atmospheric correction of ocean color data and to the quality control of seaborne, airborne, and spaceborne reflectance measurements in turbid waters are outlined.Although ocean color remote sensing has focused primarily on visible wavelengths (400-700 nm) where photosynthetic pigments have detectable absorption features, there is a growing interest in water-leaving reflectances at near infrared (NIR) wavelengths, usually taken as the range 700-1,000 nm. There are essentially three reasons for this. First, although for atmospheric correction over clear waters the NIR water-leaving reflectance can usually be taken as zero (Gordon and Wang 1994), for turbid waters it is essential to model or estimate NIR water-leaving reflectance in order to 1 Corresponding author (K.Ruddick@mumm.ac.be). 2 Present address: Polytechnic of Namibia, P/Bag 13388, Windhoek, Namibia. AcknowledgmentsThis study was funded by the Belgian Science Policy Office's STEREO program in the framework of the BELCOLOUR project SR/00/03, by the European Union under the REVAMP project EVG1-CT-2001-00049, and by PRODEX contract 15190/01. The captains, crews, and support staff of the research vessels Belgica and Zeeleeuw are thanked for their enthusiastic help with the seaborne measurements. Jean-Paul Huot and the scientists of the RE-VAMP project and the MERIS Validation Team are especially thanked for the many discussions that have helped to improve and control the quality of the seaborne measurements. Wolfgang Cordes of GKSS is acknowledged for the tests on polarization and droplet sensitivity of spectroradiometers. Rudiger Heuermann of TriOS and Jean-Pierre De Blauwe and André Pollentier of MUMM-Oostende are thanked for help with system design; Barbara Van Mol for help with figure preparation; Bouchra Nechad for discussions on the use of the near infrared (NIR) range for total suspended matter (TSM) retrieval; and Arnold Dekker for discussion of op...

The standard SeaWiFS atmospheric correction algorithm, designed for open ocean water, has been extended for use over turbid coastal and inland waters. Failure of the standard algorithm over turbid waters can be attributed to invalid assumptions of zero water-leaving radiance for the near-infrared bands at 765 and 865 nm. In the present study these assumptions are replaced by the assumptions of spatial homogeneity of the 765:865-nm ratios for aerosol reflectance and for water-leaving reflectance. These two ratios are imposed as calibration parameters after inspection of the Rayleigh-corrected reflectance scatterplot. The performance of the new algorithm is demonstrated for imagery of Belgian coastal waters and yields physically realistic water-leaving radiance spectra. A preliminary comparison with in situ radiance spectra for the Dutch Lake Markermeer shows significant improvement over the standard atmospheric correction algorithm. An analysis is made of the sensitivity of results to the choice of calibration parameters, and perspectives for application of the method to other sensors are briefly discussed.

No abstract

No abstract