Key Factors Affecting Temporal Variability in Stream Water Quality

Degraded water quality in rivers and streams can have large economic, societal and ecological impacts. Riverine water quality can vary substantially across time, so understanding the factors that influence such temporal variability is critical for designing water quality management strategies. In this study, we use a Bayesian hierarchical modelling approach to describe the spatio-temporal variability in stream water quality across multiple catchments in Victoria. We utilize monthly water quality monitoring data collected at 102 sites in the state of Victoria (South-East Australia) over 20 years. We focus on three water quality constituents: total suspended solids (TSS), nitrate-nitrite (NO_x), and electrical conductivity (EC).

We use an exhaustive search approach to identify the best set of explanatory variables of water quality temporal variability in the Bayesian hierarchical models. The fitted Bayesian hierarchical models suggest that, not surprisingly, same-day streamflow is the most important factor that controls water quality temporal variability for all constituents. Additional important predictors include soil moisture, antecedent streamflow, vegetation cover, and water temperature. The effect of the temporal predictors on water quality varies between sites, and are related to differences in catchment characteristics including topography, climate and land-use. This understanding of temporal water quality variability will be used to build a predictive spatio-temporal model. This model will be useful as an operational tool for catchment managers, to forecast stream water quality responses to potential changes such as land use and climate.