Data for salinity regressions and total dissolved solids calculations in groundwater in the United States

Groundwater salinity impacts water quality and limits usability for drinking, irrigation, and industrial purposes where salinity is high. Salinity is a measure of the mass of dissolved salts in a given volume of solution and is closely related to the concentration of Total Dissolved Solids (TDS) as well as the ability of water to conduct electricity, or specific conductance (SC) (McCleskey and others, 2025). The relationship between salinity, TDS, and SC are controlled by the major ion composition of groundwater, or hydrochemical facies. Hydrochemical facies represent the chemical reactions between groundwater and the surrounding geology and correlate with depth (Stackelberg and others, 2025). Thus, understanding the hydrochemical facies and salinity of groundwater is crucial for effective water resource management.

This data release documents groundwater data and calculations (scripted workflow written in Python) that uses the hydrochemical facies of groundwater, in conjunction with measurements of specific conductance, to accurately estimate the salinity of groundwater for the continental United States (CONUS). Additionally, TDS can be calculated several ways. TDS can be measured directly following U.S. Geological Survey (USGS) residual on evaporation method 70300 or by utilizing a summation of constituents (TDSsoc) USGS method 70301. The TDSsoc method sums the concentrations, in milligrams per liter (mg/L) for the following major ion constituents: Ca, Mg, Na, K, Cl, SO4, NO3, CO3, and SiO2. The data sources include dissolved solids data from a brackish groundwater assessment (Qi and others, 2017), produced waters geochemical data, (Blondes and others, 2023), state groundwater-quality data from the Safe Drinking Water Information System (SDWIS) contained within the workflow, and preprocessed groundwater-quality data retrieved from the National Water Quality Portal (WQP), also contained within the workflow.

An additional R script utilizes the output from the Python workflow and cleans, filters, and prepares data for geochemical modeling, ensuring that only high-quality, relevant samples are included for further analysis and facilitates the execution and output handling of PHREEQC and Pitzer calculations.