Field measurements of a and b are relatively rare in visible wavelengths and extremely scarce in UV wavelengths (but see Boss et al., this issue). The value of a is the optical sum of absorption by dissolved and particulate constituents of natural waters in combination with aw, absorption by H20. The primary contributor to absorption by dissolved constituents in natural waters is colored dissolved organic matter (CDOM);acdom is typically measured using a laboratory spectrophotometer after particles are removed from the water sample by filtration. Values for aCDOM can also be estimated from measurements of DOC concentration if DOC-specific absorption can be estimated as well. Although suspended mineral particles can sometimes make a large contribution to attenuation, especially in shallow water or near inflow from glaciers or rivers, the optically-important particles in lakes are typically phytoplankton. Spectral absorption by phytoplankton can be measured in a spectrophotometer by concentrating a water sample onto a glass fiber filter. While primarily used for visible wavelengths (Yentsch & Phinney, 1989; Mitchell 1990; Lohrenz 2000), the technique has also been used for UV wavelengths (Ayoub et al., 1996; Sosik, 1999; Helbling, et al., 1994; Belzile et al., 2002; Hargreaves 2003; Laurion et al., 2003).
Indirect measures of optical properties can be predictive of phytoplankton abundance and Kd,UV in low-CDOM systems. The concentration of the primary photosynthetic pigment in phytoplankton (chlorophyll a) can be detected in vivo by its red absorption peak (676 nm) or by fluorescence measurements (emission peak at 683 nm) in the water column. Solar-stimulated fluorescence from phytoplankton pigments can also be detected by spectral reflectance meters after correction for Raman scattering. In natural waters the cp660 signal described above primarily responds to particle concentration because of scattering at 660 nm but when the particles are predominantly biotic, cp660 is expected to covary also with absorption and also attenuation at other wavelengths. None of these indirect measures is likely to be useful alone in predicting UV attenuation over a range of depths because of photoacclimation: the deeper phytoplankton adjust to dim light by increasing the efficiency of light utilization, the concentration of chlorophyll per cell, and the absorption per unit of chlorophyll, and decreasing the proportion of UV-screening pigments (MacIntyre et al 2002). Summing Fchl and cp660, with proper adjustment of their relative contribution, might provide a useful index of changing UV attenuation with depth when direct measures of UV attenuation are unavailable.
Another optical measurement that should be related to Kd,UV in UV-transparent systems is Secchi depth (ZSD). Measurement of ZSD has been used for many years as a simple transparency index of water quality (Larson et al., 1996a). The depth at which a 20 cm white disk is barely visible under ideal conditions (flat surface, no reflections from the surface, and adequate solar radiation) depends on a combination of scattering that obscures the image of the underwater disc and absorption that diminishes the light reaching the disk from the surface. The inverse of Secchi depth (1/ZSD, unit m-1) has been shown to correlate with [Kd + c] where Kd and c are measured for the appropriate range of wavelengths dependent on the combination of human vision and peak transmission wavelengths (Tyler,1968; Preisendorfer, 1986, ) and depth-averaged from the surface to ZSD. Human visibility of black objects underwater has been shown to vary inversely with beam attenuation in green wavebands (530 nm, Davies-Colley, 1988; Zaneveld & Pegau, 2003) but the blue waveband is likely to be more important in the case of very clear water such as Crater Lake. Because phytoplankton contribute to both scattering and absorption in blue wavelengths and typically have UV-absorbing protective pigments in a high UVR environment, blue attenuation and 1/ZSD should be correlated with UV attenuation when the latter is affected by phytoplankton. In other studies where phytoplankton and suspended mineral sediments control transparency, Secchi depth has been correlated with Kd measurements for the PAR waveband and with the concentration of suspended sediments (e.g. Jassby et al., 1999). In systems where the relative contributions to optical attenuation by phytoplankton and suspended mineral particles are variable, the relationships among cp660, 1/ZSD, Kd, and phytoplankton concentration would be expected to vary somewhat, with Kd less responsive to increases in scattering than the other two measurements.
***previous*** — ***next***
