Using a GIS to interpret tritium-groundwater age distribution in Indiana
Abstract
As part of a water supply vulnerability study for the Indiana Department of Environmental Management (IDEM), several hundred groundwater samples have been tested for tritium activity, and the resulting distribution of tritium suggests complex patterns of groundwater age. To interpret the distribution of ages, a GIS tool was developed by linking the sampling locations with lithologic logs and hydrogeologic settings maps. Then, groundwater age was estimated from the tritium data using a geographically representative recharge input function and a mixing model. Separate input functions were developed for recharge in Northern, Central, and Southern Indiana. Groundwater age was estimated using the input functions with a mixing model that approximates advective dispersion and by comparing model-calculated tritium activities with groundwater tritium data. Next a preliminary analysis suggested that the hydrogeologic setting where a well is located, such as an outwash fan or a till plain, seems to be a good overall predictor for the tritium level; therefore, the statistical relationship between geologic setting and tritium levels was explored in greater detail. A statistical test of the median tritium levels suggests that the geologic setting in which a well is located does influence the observed tritium activity. Median activities for outwash and river valleys setting groups were found to be significantly greater than the median for all of the data, while medians for settings groups consisting of till plains and moraines were found to be significantly less than the global median. These results support the hypothesis that hydrogeologic setting is a good predictor of groundwater age. However, numerous wells produced anomalously young or old groundwater compared with other wells in similar hydrogeologic settings. Those anomalies were investigated using the GIS, and some wells were also re-sampled and re-analyzed for tritium. In most cases the follow-up sampling results confirm the original “anomalous” results, and furthermore, all of the wells deemed to be tritium dead with the original data, were verified to be tritium dead with the re-sampling data.
Degree
Ph.D.
Advisors
Leap, Purdue University.
Subject Area
Hydrology|Geology|Geochemistry
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.