A sensor network for high frequency estimation of water quality constituent fluxes using surrogates

Abstract

Characterizing spatial and temporal variability in the fluxes and stores of water and water borne constituents is important in understanding the mechanisms and flow paths that carry constituents to a stream and through a watershed. High frequency data collected at multiple sites can be used to more effectively quantify spatial and temporal variability in water quality constituent fluxes than through the use of low frequency water quality grab sampling. However, for many constituents (e.g., sediment and phosphorus) in-situ sensor technology does not currently exist for making high frequency measurements of constituent concentrations. In this paper we describe how water quality measures such as turbidity or specific conductance, which can be measured in-situ with high frequency, can be used as surrogates for other water quality constituents that cannot economically be measured with high frequency to provide continuous time series of water quality constituent concentrations and fluxes. We describe the observing infrastructure required to make high frequency estimates of water quality constituent fluxes based on surrogate data at multiple sites within a sensor network supporting an environmental observatory. This includes the supporting sensor, communication, data management, and data storage and processing infrastructure. We then provide a case study implementation in the Little Bear River watershed of northern Utah, USA, where a wireless sensor network has been developed for estimating total phosphorus and total suspended solids fluxes using turbidity as a surrogate.