Empirical methods for the prediction of optical properties in shallow estuaries

Date of Completion

January 2007


Physical Oceanography




Light is attenuated in aquatic systems through absorption and scattering processes caused by four components: the water itself, phytoplankton biomass, non-phytoplankton particles like minerals and detritus, and colored dissolved organic matter (CDOM). In open ocean waters, phytoplankton are the dominant attenuators of light, but in estuariee waters, all four components can significantly contribute to light attenuation. Traditionally, ecosystem models predict the diffuse attenuation coefficient with an empirical function of chlorophyll concentration (a measure of phytoplankton biomass) plus a constant background value. This method fails in shallow estuaries.^ CDOM contributes strongly to light attenuation in the shallow estuaries of southern New England. An alternative empirical prediction that includes the effects of chlorophyll, CDOM and suspended particles was developed. This empirical method can predict 85% of the variability in the diffuse attenuation coefficient of photosynthetically available radiation observed in the eleven estuaries studied, and can be evaluated using commonly measured water quality parameters. Although this more complex empirical regression should be used in sites where turbidity is substantial, the diffuse attenuation coefficient in the low turbidity estuaries of southern New England is well predicted by a simpler regression, also presented here, including only chlorophyll and CDOM.^ Colored dissolved organic matter (CDOM) is input to estuaries from their waters watersheds. CDOM concentrations were linearly inversely correlated with salinity in the shallow estuaries studied here, even in those with modest freshwater inflows, and the seaward end members were fairly uniform. The freshwater inflow to an estuary, however, is not a good predictor of the incoming CDOM concentration, even though variability exists in the CDOM concentration of the freshwater within a single estuary. No seasonal pattern exists in incoming CDOM concentration. Cross-site variability accounts for half of the observed variability in freshwater CDOM concentration. The land cover characteristics of a watershed are strongly related to the incoming CDOM concentration. An empirical regression equation, using the percent cover of four land use/land cover types m their large temporal (coniferous forest, turf, agricultural land and developed land) as the variables, predicts 45% of the observed variability in freshwater CDOM concentration in 11 southern New England estuaries. ^