146-4 Determining Groundwater Salinity Sources Using a Combined Multi-Isotope and Non-Negative Matrix Factorization Approach, Lower Valley Area of the Hueco Bolson, El Paso County, Texas

Session: Contaminants Near Groundwater-Surface Water Interfaces (Posters)

Presenting Author:

Authors:

Abstract:

Groundwater salinization is increasingly troublesome in  aquifers in and around El Paso, Texas, where concern for available water resources has grown due to aridity and increasing population. Both the groundwater and surface water have experienced an increase in total dissolved solids (TDS), along with a drop in the water table and recent decrease in annual precipitation – trends that can pose future health and economic problems for residents who primarily rely on groundwater. To address this issue, this research focuses on the Hueco Bolson Aquifer and the Rio Grande Alluvial Aquifer within the Lower Valley area of El Paso, Texas, employing a two-pronged approach to evaluate and determine solute sources and their possible end-members in groundwater. The first approach employs machine learning non-negative matrix factorization (NMF) source apportionment- a powerful multivariate statistical technique that decomposes complex geochemical datasets into distinct source profiles and their contributions while minimizing potential bias- using pre-existing public geochemical datasets and our own data. Recognizing that this approach can have difficulty in separating chemically similar sources due to limited tracer data, our second approach applies geochemical and isotopic tools  (δ¹¹B, δ²H, δ¹⁸O_H2O, ⁸⁷Sr/⁸⁶Sr,  ²³⁴U/²³⁸U,  and δ³⁴S_SO4 and δ¹⁸O_SO4) on locally collected samples to independently confirm and refine the NMF-derived end-member interpretations. We hypothesize that elevated groundwater salinity in the area results from the synergistic effects of two primary agricultural practices: flood irrigation and fertilizer application. We expect our NMF results to reveal end-member chemistries in the groundwater that could point to distinct solute contributions from both natural and anthropogenic sources, such as evaporative concentration, evaporite dissolution, and nutrient load signatures consistent with fertilizer application. Initial results have indicated strong correlations between TDS, Na, Cl, Ca, and Li. Isotopic analysis have pointed to two possible endmembers influencing groundwater mixing from ²³⁴U/²³⁸U, three from ⁸⁷Sr/⁸⁶Sr, and δ²H and δ¹⁸O_H2O results have shown that most of our samples, including our possible end-members, plot within the Rio Grande Evaporation Line and are <100 years old. This research attempts to clarify the effects of agricultural practices, groundwater-surface water interactions, and contaminant transport for the development of future water management strategies.

Geological Society of America Abstracts with Program. Vol. 57, No. 6, 2025

doi: 10.1130/abs/2025AM-9623

© Copyright 2025 The Geological Society of America (GSA), all rights reserved.