Controls on Spatial Variability in Mean Concentrations and Export Patterns of River Chemistry Across the Australian Continent

• Published in: Water resources research Vol. 58; no. 12
• Main Authors: Liu, Shuci, Dupas, Rémi, Guo, Danlu, Lintern, Anna, Minaudo, Camille, Bende‐Michl, Ulrike, Zhang, Kefeng, Duvert, Clément
• Format: Journal Article
• Language: English
• Published: 01.12.2022
• Subjects: Bayesian hierarchical modeling; Bayesian model averaging; catchment characteristics; concentration‐discharge relationship; spatial variability; water quality
• ISSN: 0043-1397; 1944-7973
• DOI: 10.1029/2022WR032365

The state and dynamics of river chemistry are influenced by both anthropogenic and natural catchment characteristics. However, understanding key controls on catchment mean concentrations and export patterns comprehensively across a wide range of climate zones is still lacking, as most of this research is focused on temperate regions. In this study, we investigate the catchment controls on mean concentrations and export patterns (concentration–discharge relationship, C–Q slope) of river chemistry, using a long‐term data set of up to 507 sites spanning five climate zones (i.e., arid, Mediterranean, temperate, subtropical, tropical) across the Australian continent. We use Bayesian model averaging (BMA) and hierarchical modeling (BHM) approaches to predict the mean concentrations and export patterns and compare the relative importance of 26 catchment characteristics (e.g., topography, climate, land use, land cover, soil properties and hydrology). Our results demonstrate that mean concentrations result from the interaction of catchment indicators and anthropogenic factors (i.e., land use, topography and soil), while export patterns are influenced by topography. We also found that incorporating the effects of climate zones in a BHM framework improved the predictability of both mean concentrations and C–Q slopes, suggesting the importance of climatic controls on hydrological and biogeochemical processes. Our study provides insights into the contrasting effects of catchment controls across different climate zones. Investigating those controls can inform sustainable water quality management strategies that consider the potential changes in river chemistry state and export behavior. Plain Language Summary Riverine water quality can change markedly between locations. Understanding the underlying causes of these differences is of great importance for water quality management. In this study, we analyze water quality monitoring data from 507 catchments across the Australian continent, spanning five major climate zones, using a statistical modeling approach. We identify the key catchment characteristics influencing water quality differences between rivers and develop models that predict these differences. We find that land use, topography and soil properties have the highest impact on water quality, while topography tends to control the patterns of solute export. This work provides data‐driven evidence for catchment managers, which can help them develop effective water quality management strategies. Key Points Consideration of climate zones in a hierarchical modeling structure improves the predictability of both mean concentrations and C–Q slopes Land use, topography and soil are the most influential factors for mean concentrations, while topography controls export patterns The influence of catchment controls on mean concentrations/C–Q slopes varies across climates zones