Decomposing the Bulk Electrical Conductivity of Streamflow To Recover Individual Solute Concentrations at High Frequency

The ability to evaluate stream hydrochemistry is often constrained by the capacity to sample streamwater at an adequate frequency. While technology is no longer a limiting factor, costs and sample management can still be a barrier to high-resolution water quality instrumentation. We propose a new framework for investigating the electrical conductivity (EC) of streamwater, which can be measured continuously through inexpensive sensors. We show that EC embeds information about individual ion content that can be isolated to retrieve solute concentrations at high resolution. The essence of the approach is the decomposition of the EC signal into its "harmonics", i.e., the specific contributions of the major ions that conduct current in water. The ion contribution is used to explore water quality patterns and to develop algorithms that reconstruct solute concentrations starting from EC during periods where solute measurements are not available. The approach is validated on a hydrochemical data set from Plynlimon, Wales, showing that improved estimates of high-frequency solute dynamics can easily be achieved. Our results support the installation of EC probes to complement water quality campaigns and suggest that the potential of EC measurements in rivers is currently far from being fully exploited.