Official websites use .gov
A .gov website belongs to an official government organization in the United States.
Secure .gov websites use HTTPS
A lock () or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.
A new 3-D model predicts pH in groundwater at all depths across the 25-state span of the glacial aquifer system, reports an article by the USGS. The glacial aquifer system provides more water—about 2.6 billion gallons per day—for domestic and public supplies than any other aquifer in the United States.
Many trace elements, such as manganese, can be released from aquifer sediment into groundwater under acidic (low pH) conditions. Acidic groundwater also contributes to corrosivity. The most acidic groundwater in the glacial aquifer system was predicted to occur primarily in the northeast. Other trace elements, such as arsenic, can be released into groundwater under alkaline (high pH) conditions. Alkaline groundwater was predicted to occur primarily in the upper Midwest and corresponds to areas with groundwater known to have elevated arsenic concentrations.
The results of the “machine-learning” model indicate that the carbonate content of aquifer materials strongly controls pH; the highest pH conditions occur in aquifers with relatively high carbonate content and with long water-sediment contact times and distances. Conversely, groundwater moving through thin sediments without a carbonate-rock component have more acidic pH conditions. The pH of groundwater generally increases as it flows through the aquifer system.
The results of the modeling demonstrate the usefulness of machine-learning methods to predict water-quality conditions in a large, complex aquifer system. Such predictions can be used for the design of future monitoring programs, to aid well owners in determining whether their water should be tested for high arsenic or high manganese concentrations, or to identify areas where groundwater is likely to be corrosive. The pH predictions also may be useful as input to other models that predict where water-quality constituents such as manganese or arsenic may be elevated. This study is the first of three that use machine-learning methods to estimate groundwater-quality conditions throughout the glacial aquifer.
