Report PDF (12.4 MB)

Abstract

Understanding of karst flow systems can be complicated by the presence of solution-enlarged conduits, which can transmit large volumes of water through the aquifer rapidly. If the geochemistry at a well can be related to streamflow or spring discharge (springflow), or both, the relations can indicate the presence of recent recharge in water at the well, which in turn might indicate that the well intersects a conduit (and thus a major flow path). Increasing knowledge of the occurrence and distribution of conduits in the aquifer can contribute to better understanding of aquifer framework and function. To that end, 26 wells in the Barton Springs segment of the Edwards aquifer, Austin, Texas, were investigated for potential intersection with conduits; 26 years of arbitrarily timed specific conductance measurements in the wells were compared to streamflow in five creeks that provide recharge to the aquifer and were compared to aquifer flow conditions as indicated by Barton Springs discharge. A nonparametric statistical test (Spearman’s rho) was used to divide the 26 wells into four groups on the basis of correlation of specific conductance of well water to streamflow or spring discharge, or both. Potential relations between conduit intersection by wells and ground-water geochemistry were investigated through analysis of historical major ion and nitrate geochemistry for wells in each of the four groups. Specific conductance at nine wells was negatively correlated with both streamflow and spring discharge, or streamflow only. These correlations were interpreted as evidence of an influx of surface-water recharge during periods of high streamflow and the influence at the wells of water from a large, upgradient part of the aquifer; and further interpreted as indicating that four wells intersect major aquifer flow paths and five wells intersect minor aquifer flow paths (short, tributary conduits). Specific conductance at six wells was positively correlated with spring discharge, which was interpreted as not intersecting a flow path (conduit). Of the 11 wells for which specific conductance did not correlate with either streamflow or spring discharge, no interpretations regarding flow-path intersection by wells were made. In some cases, specific conductance data might not have indicated intersection with a flow path because of small sample sets. Water in the Barton Springs segment generally is a calcium-magnesium-bicarbonate type, although some water compositions deviate from this. Multiple geochemical processes were identified that might affect geochemistry at the wells, but in general the geochemical composition of ground water, except for dilution by surface-water recharge, was not related to intersection of a well with a flow path. Some samples from wells indicate inflow of water from the saline zone to the east; this inflow is associated with low streamflow and spring discharge. Other samples indicate that the aquifer at some wells might be receiving water that has been in contact with rocks of the Trinity aquifer; this mixing is most evident when spring discharge is high. Occurrence of nitrate in ground water was unrelated to intersection of flow paths by wells and appeared to be the result of localized contamination. However, most of the wells with one or more samples contaminated by nitrate are in the more densely populated parts of the study area.