Turbidity and temperature patterns in a reservoir and its primary tributary from robotic monitoring: Implications for managing the quality of withdrawals

Abstract

Abstract A robotic water quality monitoring network, consisting of 2 in-reservoir profiling platforms and a unit positioned in the primary tributary, was used to document patterns of temperature (T) and turbidity (Tn) and the impact of runoff events on Tn levels over the spring to fall interval of 3 years. The patterns were evaluated in the context of water quality goals for T and Tn for reservoir withdrawal and potential benefits of specified management alternatives. Exceedences of a proposed withdrawal T goal of 21.1 °C (70 °F) and a conservative representation of the Tn goal of 15 NTU were demonstrated. The exceedences occurred in late summer during intervals of extensive drawdown for T and irregularly following runoff events for Tn. Conspicuous increases in Tn in the tributary and lacustrine portions of the reservoir were demonstrated in response to all monitored runoff events (n = 34). Lacustrine levels exceeded 75 NTU following several major events. The turbidity load from the tributary usually enters the reservoir as a turbid density current during runoff events, at a depth predicted well by stream temperature. The high frequency and vertical resolution attributes of the robotic monitoring (2 profiles per day for reservoir deployments) were invaluable in resolving impacts because these were usually attenuated within a week following events. The diminishment of reservoir Tn levels following events was demonstrated to be well represented by a first-order loss rate. The loss rate for the largest runoff event of the study (return interval ~25 y) was more than an order of magnitude lower than any other, indicating a difference in settling attributes of turbidity-causing particles for this extreme case. Benefits from a shift to delayed withdrawals for the existing intake, or installation of a multi-level intake facility, for improving status with respect to the water quality goals are suggested from the monitoring data. However, continuously meeting the Tn goals for particularly severe events may not be feasible through these approaches.