The attenuation of solar UV radiation in lakes and the role of dissolved organic carbon

TLDR

Diffuse attenuation coefficients for UV radiation were measured in the mixed layers of 65 lakes across Alaska, Colorado, Pennsylvania, and Argentina, and a multivariate model was built using optical and chemical variables to predict Kd. The study found that DOC concentration explains 87–96 % of the variation in UV attenuation, with a univariate DOC‑based power model providing the best estimates of Kd across the UV‑A and UV‑B ranges, enabling UV transparency predictions from archival DOC or color data.

Abstract

Diffuse attenuation coefficients ( K d ) for solar UV radiation (UVR) (305, 320, 340, 380 nm, and PAR) were measured in the mixed layer of 65 lake sites in Alaska, Colorado, and Pennsylvania and the Bariloche region of Argentina. Integrated mixed layer samples of lake water were concurrently collected, and a multivariate approach was used to model K d with a number of optical and chemical variables. Substantial variation in transparency was observed among lakes. Attenuation depths ( z 1% ) for UV‐B radiation ranged from several centimeters to > 10 m. In some shallow, low DOC (dissolved organic C) lakes typical of high elevation ecosystems, substantial fluxes of UVR penetrated the entire water column. In deeper lakes with low DOC concentrations, high fluxes of UVR were found in a significant proportion of the mixed layer. Much of the among‐lake variation in K d (87–96%) was explained by differences in DOC concentration, which strongly influenced dissolved absorbance. On average, dissolved absorbance accounted for between 33% (for PAR) and 68% (for 305 nm) of K d measured in situ. Throughout the solar UV‐A and UV‐B range, K d was best estimated with a univariate power model based solely on DOC concentration. Models are also presented that relate absorption coefficients to K d . These models can be used with archival DOC or color data to provide approximate estimates of UV transparency of lakes.