Predicting light penetration into river waters

Lighting in rivers often needs to be quantified, particularly for modeling benthic plant growth, but is seldom measured because of difficulties associated with limited depth and strong currents. Therefore, methods for predicting light attenuation from river water quality data would be very useful. We used measurements of the diffuse light attenuation coefficient, K-d (m(-1)), at 17 optically diverse rivers in New Zealand to develop simple empirical models of light penetration as functions of the beam attenuation coefficient at 550 nm, c(550) (m(-1), an index of visual water clarity) and the light absorption coefficient of membrane filtrates at 340 nm, g(340) (m(-1), an index of colored dissolved organic matter). The beam attenuation coefficient can be measured by beam transmissometer or estimated, as in this study, from black disc visibility observations. Alternatively, nephelometric turbidity, T-n (an index of light scattering), which is more commonly measured in water quality monitoring programs, can be used to predict Kd. The models performed satisfactorily when tested over a wide range of optical water quality (varying with flow) at one river site. We expect that these empirical models will have wide practical application for estimating light availability in rivers and streams.