Human health risk from manganese in groundwater
Machine learning is the key to uncovering where populations across the globe are at risk from manganese in their groundwater-sourced drinking water.
EXPLORE THE ARCHIVE
Publications and source records attributed to Katherine J. Knierim.
Machine learning is the key to uncovering where populations across the globe are at risk from manganese in their groundwater-sourced drinking water.
Groundwater is increasingly needed for water supplies but may have limited utility in some locations because of its salinity. Salinity, often expressed as total dissolved solid (TDS), is frequently estimated using specific conductance (SC) measurements. However, the commonly used proxy (0.65 multiplied by SC to indicate TDS, common in many handheld meters) can result in inaccurate TDS estimates. First, the TDS–SC relationship is not linear over the entire concentration range of groundwater. Furthermore, the TDS (and salinity)–SC relationships vary substantially depending on the major-ion composition. Here we develop a proxy method utilizing SC and major-ion water type to estimate TDS and salinity specifically for groundwaters. Compared to most surface waters, groundwater tends to have a wider range of salinity (fresh to highly saline) and higher concentrations of bedrock-derived solutes such as carbonate ions, silica, and many other ions. The dataset used to develop the proxies includes water chemistry data from 149,059 discrete groundwater samples. The groundwater proxies, which employ nonlinear log–log relations, utilize five water types (HCO 3 , Cl, Ca-Mg-SO 4 , Na-K-SO 4 , and mixed waters), are accurate (median percent difference between TDS and salinity determined using the proxy compared to discrete measurements was <±0.8%) over a wide range of SC (up to 200 mS/cm), rapid, cost-effective, and can be measured on-site.
This study characterized groundwater age across the Mississippi River Valley alluvial aquifer (MRVA). Groundwater samples from 69 MRVA wells and 19 wells in Tertiary units of the Mississippi embayment aquifer system (MEAS) were analyzed for sulfur hexafluoride (SF 6 ), tritium ( 3 H), helium (He), and (or) carbon-14 of dissolved inorganic carbon ( 14 C). The age distributions of 89 samples were estimated by fitting lumped parameter models to processed tracer concentrations with the U.S. Geological Survey software TracerLPM. Mean ages of MRVA groundwater samples ranged from 12 to 22,000 years, with a median of 140 years. Mean ages of MEAS groundwater samples ranged from 230 to 52,000 or more years, with a median of 13,500 years. The spatial distribution of MRVA groundwater ages was found to be influenced by depth, inflow of groundwater from deeper units, and soil saturated hydraulic conductivity. In parts of the MRVA, the spatial distribution of MRVA groundwater ages was found to be influenced by annual recharge and (or) annual groundwater pumpage.
Global demand for lithium, the primary component of lithium-ion batteries, greatly exceeds known supplies, and this imbalance is expected to increase as the world transitions away from fossil fuel energy sources. High concentrations of lithium in brines have been observed in the Smackover Formation in southern Arkansas (>400 milligrams per liter). We used published and newly collected brine lithium concentration data to train a random forest machine-learning model using geologic, geochemical, and temperature explanatory variables and create a map of predicted lithium concentrations in Smackover Formation brines across southern Arkansas. Using these predicted lithium maps with reservoir parameters and geologic information, we calculated that there are 5.1 to 19 million tons of lithium in Smackover Formation brines in southern Arkansas, which represents 35 to 136% of the current US lithium resource estimate. Based on these calculations, in 2022, 5000 tons of dissolved lithium were brought to the surface within brines as waste streams of the oil, gas, and bromine industries.