Travertine deposits as records of groundwater evolution in urbanizing environments

Urban development has multiple impacts on groundwater, including changes in patterns of recharge, alteration of groundwater flow paths, and increased water contamination. For the most part, these impacts have only recently begun to be appreciated, and in most aquifers there is a lack of data available on recharge and water quality prior to urbanization for comparison. This lack of data across the period of urbanization makes the prediction of the impacts of future development difficult. I propose that recently deposited spring water calcite (travertine) may provide a proxy record of changing groundwater composition over time. The city of Austin, Texas has undergone a large increase in urbanization over the past 40 years. Because the chemistry of municipal water in Austin is distinctive from the local groundwater, I have examined the possibility that the amount of urban water input to local groundwater recharge can be assessed using the isotopic composition of strontium (⁸⁷Sr/⁸⁶Sr) in combination with elemental concentrations and ratios of Mg/Ca and Sr/Ca. This study examines the current depositional environment of three travertine-depositing springs in the Austin area. It makes use of the modern system, including experimental growth of calcite on artificial substrates in the springs, to interpret a 30-year temporal record provided by one of the three studied springs. Three travertine-depositing springs were studied over the course of one year to examine water chemistry and strontium isotope composition changes on a seasonal basis. Elemental concentrations and strontium isotope ratios remained fairly constant throughout the study year, and showed the influence of urban water in the isotopic compositions of the spring water. Glass plate substrates were employed over a fourmonth time period, to examine monthly changes in travertine growth and isotopic and elemental compositions. Using these substrate experiments, I determine travertine growth rates and effective trace element partition coefficients. These partition coefficients are higher than values determined from laboratory experiments from the literature, indicating that partitioning behavior in natural systems may be different than that of controlled laboratory experiments. Applying these partition coefficients to the travertine temporal record provides a more accurate interpretation of past water compositions. Comparison of stable isotopes of carbon and oxygen between spring waters and plate calcites revealed that the plate carbonates (travertines) are not forming in isotopic equilibrium with their dripwaters. Based on this, interpretation of stable isotope records from these travertines should be treated with caution. Isotopic and elemental analyses were conducted on a spring-fed travertine that preserved 30 years of deposition on a roadcut through Lower Cretaceous marine limestones. The spring is actively precipitating calcite, and appears to have grown continually continually since the excavation of the roadcut. The contributing area to this spring has undergone an increase in urban land use of approximately 40% since 1983. ⁸⁷Sr/⁸⁶Sr data for the travertine and spring waters show a small but resolvable increase over time, from 0.70790 to 0.70797. These values for ⁸⁷Sr/⁸⁶Sr are higher than those for Lower Cretaceous marine limestones, and lower than values for municipal water as well as soil exchangeable Sr. The increase in ⁸⁷Sr/⁸⁶Sr may be accounted for in three ways: 1) Through an increased relative proportion of municipal water input to the groundwater in the spring’s contributing area; 2) Through increased overall recharge in the area, resulting in more water-soil interaction; or 3) A combination of the two. Modeling of karst flow processes (water-limestone interaction and fluid mixing) shows that the trend in elemental ratios and ⁸⁷Sr/⁸⁶Sr may be accounted for by processes of soil water-limestone interaction and mixing with urban waters. Models also indicate that this fluid mixing likely occurs in the epikarst zone, and that urban water may account for as much as 80% of the discharge from the spring. Values for ⁸⁷Sr/⁸⁶Sr in spring waters are likely dampened by limestone interaction during times of low recharge. This model of waterlimestone interaction and urban water mixing implies that urbanization in this particular area resulted in increased recharge and that this recharge may be quantifiable over time. Travertine proxies represent a previously unused resource for examining urban recharge in karst terrains.