Estimating primary production in coastal waters with remote sensing

VI ABSTRACT Remote sensing of ocean color is based on the measurement of water-leaving radiance in visible part of the electromagnetic spectrum. The total signal detected by the satellite sensor is corrected for the atmospheric noise. Then the concentrations of phytoplankton pigments in the surface layers of the ocean can be derived by using empirical algorithms. The algorithms in use generally fail in coastal waters, which are optically complex due to the presence of other substances like suspended solids. There are efforts on the development of new algorithms and new sensors to infer concentrations of phytoplankton pigments in coastal waters. The NOAA's Advanced Very High Resolution Radiometer (AVHRR) was designed for meteorological purposes, not for use as an ocean color sensor. The most successful use of AVHRR data in oceanography is the determination of sea surface temperature maps. Although it does not have the sensitivity in the red and near- infrared to examine open ocean waters (case-1 waters), its sensitivity is suitable to examine the patterns in water clarity at turbid estuaries (case-2 waters) (Stumpf, 1987, 1992, Stumpf & Frayer 1997, Stumpf & Pennock 1989, 1991, Stumpf & Tyler, 1988, Tyler & Stumpf, 1989, Stumpf et al, 1999, Woodruff et ah, 1999). Eutrophication due to the supply of organic matter is one of the most important environmental problems of Izmir Bay. Monitoring the levels of chlorophyll - as an indicator for eutrophication - through remote sensing has substantial benefits for the management of eutrophication.VII The objective of this study is to construct a model, which estimates the distribution of aquatic chlorophyll-a concentrations in the surface layers of the Izmir Bay with interpretations of remote sensing observations and calculation of primary production based on empirical algorithms. The dataset covers in situ measurements carried out in Izmir Bay for the period of 1996-1998 and corresponding AVHRR (NOAA-14) images. Stumpf s (1992) algorithm is used to remove atmospheric effects from the images and retrieve the water leaving reflectances. Then a model is constructed to calibrate those reflectances with the in situ chlorophyll-a and attenuation coefficient data. Finally an empirical model is applied to estimate primary production from the remotely retrieved chlorophyll-a data.