High-Frequency Monitoring System
We monitored high-frequency light absorbance at multiple depths in FCR using a s::can Spectrolyser UV-Visible spectrophotometer (s::can Messtechnik GmbH, Vienna, Austria). This spectrophotometer was coupled with a multiplexor pumping system (’MUX’ from MultiplexÔ, LLC; for technical details on the multiplexor pumping system and the sensor setup, refer to Birgand et al. 2016 and Figures S1-S3). The MUX pumps water samples from multiple depths into a flow-through cuvette where the UV-vis absorbance spectra of the sample are measured by the spectrophotometer. The system used in our study collected measurements of light absorbance every 2.5 nm wavelengths from 200 nm to 732.5 nm (optical path length of 10 mm) approximately at an hourly time step for six monitoring depths in the reservoir.
The MUX system was used to collect high-frequency data during two time periods: reservoir turnover (“Turnover Deployment”) and the initiation of HOx operation ( “Oxygen On Deployment”). The Turnover Deployment captured the natural oxygenation and mixing processes that occurred during reservoir turnover and lasted from 16 October to 9 November 2020. In this study, fall turnover was defined as the first time when the temperature differential between 0.1 m and 9 m depths in the reservoir was <1 oC after summer stratification (following McClure et al. 2018), which occurred on 2 November 2020. During this time period the HOx system was on, so the hypolimnion was oxic before turnover, but the reservoir was thermally stratified. The Oxygen On Deployment was conducted between 26 May and 21 June 2021; during that time the HOx system was initiated on 11 June 2021 at 11:00 EDT. This deployment captured the engineered oxygenation and mixing processes resulting from the initiation of HOx operation. The reservoir was thermally-stratified and the hypolimnion was anoxic (DO < 1 mg/L) prior to HOx operation and thus, while HOx operation added oxygen to the hypolimnion, we observed a limited increase in DO concentrations due to high chemical oxygen demand. The HOx system induced internal mixing within the hypolimnion, but the overall thermal stratification of the reservoir was not affected.
We took multiple steps to limit the influence of fouling of the internal components of the MUX system, due to precipitation of Fe in contact with oxygen in the measuring cuvette. Between each pump cycle, deionized water was flushed through the system. At the end of each pump sequence (one sample from each depth), dilute hydrochloric acid (5%) was automatically pumped through the system and allowed to sit in the flow-through cuvette for approximately 2 minutes. We also collected a reference measurement in air at the end of each cycle, which was useful in determining the extent of fouling. Despite these efforts, some fouling was still evident during certain time periods (see Figures S4-5). Fouling was most pronounced in the lower wavelengths (200-250 nm; see Figures S6-7) and therefore we removed values for wavelengths less than 250 nm before fitting PLSR models.