Ultraviolet Radiation and Bio-optics in Crater Lake, Oregon, 2005
Other optical and bio-optical measurements
Bio-optical signals (most of which were described in Larson et al., 1996a) were recorded by a SeaTech transmissometer (25 cm 660 nm) and Wetlabs, Inc. WetStar chlorophyll-a fluorometer integrated with a SeaBird CTD profiler. Beam attenuation c660 was calculated from 2 m binned transmittance data (c660= -Ln(T)*100/25), then converted to cp660 by subtracting cw660, the beam attenuation coefficient for pure water (0.411 from cw = aw + bw; aw from Pope & Fry 1997; bw from Morel 1974). Previous estimates of cw660 (Smith & Baker, 1981; Zaneveld & Bartz 1984; Bishop 1986) used to calculate cp660 have been superseded by much-improved measurements of aw (Pope & Fry 1997).
Particles and whole water were analyzed on several occasions for organic carbon content. On two dates in 1999 (corresponding to cp660 profiles) samples were collected from three depths in 30 L. Niskin samplers (acid-washed), transferred to 20 L acid-washed carboys and 4–8 liters were filtered through pre-combusted GF/F filters. Filters were analyzed on a Carlo Erba NA1500 Carbon/Nitrogen/Sulfur Analyzer using GF/F filter blanks and cystine standards. Whole water samples were also collected similarly (Urbach et al., 2001) on three dates from 12 depths and analyzed for total organic carbon (high temperature combustion using Shimadzu TOC-5000). On two dates the measurements of particulate organic carbon (POC g C/m3) at three depths were correlated with cp660 (binned at 2 m intervals) yielding an average relationship wherecp660=0.019*[POC uM C] (r2=0.53). Because of the small number of samples and low r2 value we used the relationship reported with greater precision by Boss et al. (this issue) where cp660=0.032*[POC uM C]-0.024 (r2=0.996) to calculate POC for all depths from cp660. We adjusted the offset term each month for slight variations in transmissometer baseline in order to match the deep particulate signals to 0.38 ?M C for depths 300–500m in all summer months. On 20 August 2001 spectral absorption of particulate samples was analyzed. Water (500–1000 ml) from five depths was filtered on GF/F filters (Whatman, with nominal retention of diameters greater than about 0.7 ?m). Optical density was measured with the filter attached to a quartz disk in a Shimadzu 1601UV spectrophotometer using a modified Quantitative Filter Technique (QFT, Yentch & Phinney 1989; Mitchell, 1990) adapted for UV wavelengths (Helbling, et al., 1994; Ayoub et al., 1996; Sosik, 1999) and corrected for pathlength amplification using the method of Lohrenz (2000).
Filter photometer depth profiles began with Utterback et al., (1942), using a custom-built underwater photometer in July 1940 (they employed Schott BG12 filters, URL http://www.us.schott.com/, and a Weston cell, described by Barnard, 1938; the combined response curve has a peak at 450 nm with half-maximum responses at 390 and 490 nm). During 1968–1991 several Kahl filter photometers with similar properties to the instrument described above (Kahlisco, Inc. underwater and deck sensors equipped with clear, red, green, and blue filters over a photodetector cell, see Larson, 1972 for added details) were used to collect light profiles to a typical depth of 145 m. The blue filter data were converted to Kd,blue by calculating transmittance (Ed,z/Ed,o) from raw deck (Edo) and underwater (Edz) data and then converting this using Kd,blue = Ln(Ed,o/Ed,z)/(Z2-Z1). Kd,blue was then averaged from 10 m to the Secchi depth (40 m if no Secchi depth). Irregular photometer Kd,blue values near the surface were excluded as needed.