Abstract
Identification of the key variables that influence spatial variation in stream water quality is crucial for designing sustainable water management strategies. In this study, we investigated the key variables that influence the spatial variability of stream of water quality, across multiple watersheds. This study uses water quality data collected over 19 years for 59 watersheds located in the Southeast Atlantic region of the United States, which includes the states of North Carolina, South Carolina, and Georgia. A conceptual modeling framework was developed to understand the linkage between the long-term mean water quality constituents (Total nitrogen, Total phosphorus, Turbidity, and pH) and the watershed characteristics (e.g., topography, land use/cover, soil type), streamflow data, and climatic variables (precipitation and temperature). The modeling results suggest that not only anthropogenic variables influence the mean water quality constituents, but other watershed characteristics, such as soil properties, have a significant impact. The natural watershed characteristics explain most of the spatial variability in the mean Turbidity and pH values in streams. The modeling results also suggest that once land use and soil properties are considered, watershed topography has a limited role to explain the variation in the mean water quality. Overall. the developed watershed models can be used to forecast stream water-quality responses to future land use, climate, soil, and land management changes within the study area. (C) 2020 Elsevier B.V. All rights reserved.