Surrogate Models for Total Dissolved Solids and Major Ion Concentrations at Real-Time Water-Quality Sites, Arkansas River Basin, Colorado, October 1989 - December 2023

From 2022 through 2024, the U.S. Geological Survey (USGS) conducted a study in cooperation with the Arkansas River Basin Regional Resources Planning Group (Southeastern Colorado Water Conservancy District, Colorado Springs Utilities, Pueblo Water, Lower Arkansas Valley Water Conservancy District, Aurora Water, and Upper Arkansas Water Conservancy District) to establish a basin-wide water-quality monitoring network by leveraging ongoing real-time measurements of streamflow and specific conductance (SC) and discrete water-quality data at 13 sites along the main stem Arkansas River and Fountain Creek from the USGS site 07108600 Arkansas River at Granite, Colorado, to the USGS site 07137500 Arkansas River near Coolidge, Kansas.

In 2021, the USGS installed a new water-quality monitor at USGS site 07099971 Arkansas River below Moffat Street, Colo. (Ark blw Moffat), to replace the existing water-quality monitor at USGS site 07099970 Arkansas River at Moffat Street, Colo. (Ark Moffat). Water-quality monitors at Ark Moffat and Ark blw Moffat were run concurrently from May 2021 through December 2023, prior to the Ark Moffat site being discontinued in December 2023. This was done to develop a relation between SC values measured at the two sites to facilitate use of the historical SC data collected at Ark Moffat from October 1988 through December 2023.

Daily mean streamflow data were compiled from the USGS National Water Information System (NWIS) database and the Colorado Division of Water Resources (DWR) Colorado's Decision Support Systems (CDSS) database. Discrete water-quality data were compiled from the USGS NWIS database. These data were used to develop surrogate models describing the relations between SC values and calcium (Ca), magnesium (Mg), sodium (Na), sulfate (SO4), total dissolved solids (TDS), and streamflow values at the 13 sites. Data from October 1989 through September 2023 were used in the analysis.

This data release provides the metadata, input, output, and script files and describes the process steps used to develop: (1) linear regression relations between discrete SC values and Ca, Mg, Na, SO4, and TDS concentrations, (2) multiple linear regression relations between discrete SC values and daily mean streamflow values to compute SC attributed to streamflow at each site, and (3) a multiple linear regression relation between SC values measured at Ark Moffat and Ark blw Moffat as a function of streamflow measured at Ark Moffat. Surrogate models (that is the linear and multiple linear regression relation equations) developed as part of this study can be used to compute real-time Ca, Mg, Na, SO4, and TDS concentrations and flow-adjusted SC and TDS values using real-time measurements of SC and streamflow at the 13 sites to gain a better understanding of potential changes in water quality related to changes in water use and water management.