Controls of space-time variance of water chemistry in river networks

River water quality is essential for ecosystem function and human well-being, yet anthropogenic impacts, such as pollutant input from agricultural activities or waste water, threaten water resources. An effective design of water quality monitoring networks is crucial to understanding and mitigating these impacts. However, optimizing monitoring is challenging because of the spatial and temporal variability of water quality, i.e. solute concentrations, driven by landscape and hydroclimatic heterogeneity.

This study uses a stochastic modeling approach applied to artificial river networks to explore how landscape and hydroclimatic heterogeneity at different spatial scales shape the space-time variance of water chemistry. Building on a previously developed headwater-scale stochastic water quality model, we simulated daily discharge and solute concentration time series for equal area subcatchments within these networks. We systematically varied the spatial configuration of subcatchment solute source concentration across the network, the source zone distribution within subcatchments, and imposed different hydroclimatic regimes. Simulated discharge and solute loads were routed through the network, incorporating in-stream processing, to generate water quantity and quality time series for each network node. A global sensitivity analysis using the Morris method was performed to assess the influence of key parameters on the space-time variance of solute concentration.

The results of the sensitivity analysis revealed that the macro-scale landscape configuration of source concentrations controls the spatial variability of solute concentrations in rivers and spatial stability, i.e. the persistence of spatial patterns through time. The relative influence of structured and random landscape heterogeneity on spatial variability was scale dependent, with distinct patterns observed across different stream orders. In contrast, subcatchment-scale processes, such as the source zone distribution, and the hydroclimatic forcing regulate temporal variability of water quality and synchrony between subcatchments. We conclude that optimal water quality monitoring network design should thus quantify spatial and temporal variability across scales, leveraging concepts like spatial stability and synchrony to maximize information gained and explicitly accounting for multiscale landscape heterogeneity.