Dissolved oxygen systematics in the Tucson Basin Aquifer, Arizona

Abstract

Dissolved oxygen concentrations in ground water were measured at 46 locations within the Tucson basin using the "modified Winkler" titration method. Ground water sampled was predominantly representative of the upper 150 meters of unconfined Pleistocene siliclastic alluvium (the Fort Lowell Formation). Oxygen concentrations ranged between 1.3 and 7.9 (+/-0.5) mg/1 and did not vary significantly with respect to when the well was sampled. These findings along with previous work suggest that valley-fill aquifers within the Basin and Range are, as a rule, "oxic" geochemical environments. Dissolved oxygen concentrations vary systematically with respect to residence time and distance from the recharge area. The lowest oxygen concentrations (1.3-4.5 mg/1) were associated with "prototype" recharge (very young, cool, dilute, and high pCO₂) indicating that oxygen loss occurs during, and is virtually limited to, an early stage of chemical evolution. Oxygen removal is probably facilitated by the elevated water table below the influent Tanque Verde Creek which supports a relatively dense strand of phreatophytes - a potential source of organic detritus. Mass balance calculations suggest that oxidation of biotite within Recent stream gravel can also result in oxygen removal. The negative oxygen flux associated with recharge is not sustained; thus Tucson basin ground water remains aerobic. Transverse dispersive influx from the vadose atmosphere represents the inferred mechanism of oxygen-transport. This is facilitated by relatively high ground-water velocity immediately downgradient from the recharge area. Transverse dispersion rapidly oxygenates slightly older ground water and as a result, oxygen-saturated conditions are prevalent throughout much of the interior basin. Lower oxygen concentrations (approximately 3.5 mg/1) are associated with an older, warmer (presumably deeper), sodium-bicarbonatesulfate facies. Oxygen removal here is assumed to result from the oxidation of residual organic matter and reduced transversé dispersive influx. Measured electrode potentials were all positive but did not vary systematically with respect to ground-water flow nor were they thermodynamically related to (0₂)/(H₂0) Nernst potentials.