3.9 Reconstructing Contamination History

When combined with measurements of contaminant concentrations, groundwater age data can be used to reconstruct contaminant histories. One early study took place in the early 1990s and reconstructed the history of agricultural contamination from measurements of both CFCs-12 and nitrate in groundwater (Böhlke and Denver, 1995). The Locust Grove catchment in Maryland, USA (Figure 25) is dominated by agricultural land uses, and discharges to Chesapeake Bay. CFC-12 dating was used to estimate the mean rate of groundwater recharge to the catchment, and the age distribution within the groundwater and in groundwater discharging to streams. Groundwater ages ranged from pre-1940 (background CFC concentrations) to modern, while ages of groundwater discharging to streams ranged from pre-1940 to approximately 1975. CFC-11 also was measured, but was found to be subject to microbial degradation (Section 5.1), so was less useful as a dating tool. Some measurements of 3H and 3He were made to confirm groundwater ages obtained with CFC-12.

Many groundwater samples from within the catchment have nitrate (NO3) concentrations above 700 mol/L (10 mg/L-N). Although most of the streams within the catchment receive groundwater, nitrate concentrations in streams are lower than those observed in groundwater, with concentrations in the two main tributaries between approximately 140 and 700 mol/L. The high nitrate concentrations in both groundwater and surface water are attributed to agricultural contamination. Some groundwater samples have lower nitrate concentrations, and these values are attributed to infiltration prior to extensive use of nitrogen fertilizers and/or to denitrification of NO3 by microbes in the subsurface. The authors found a very strong correlation between nitrate concentrations and groundwater age (as determined with CFC-12), with relatively low concentrations (< 500 mol/L) in water recharged between 1940 and late 1960s, but with increasing concentrations between 1970 and 1990 (Figure 42). The relationship mirrors that of regional fertilizer use (Böhlke and Denver, 1995).

The nitrate flux into streams in the region at the time of this study was less than the nitrate flux into groundwater, in part because of denitrification, but also because groundwater discharge to streams is decades old (and so has lower nitrate concentrations than more recent groundwater recharge). Over time, nitrate fluxes to streams are expected to increase.

Graph showing NO<sub>3</sub> concentration in groundwater recharge as a function CFC-12 age of groundwater samples
Figure 42 – NO3 concentration in groundwater recharge as a function CFC-12 age of groundwater samples. The observed relationship shows good agreement with the pattern of regional use of nitrogen fertilizers (After Böhlke and Denver, 1995).

The potential of groundwater dating to assist in investigations of point-source contamination of groundwater is illustrated by a study from Cape Cod, Massachusetts (Figure 18). The site is on a military reservation, and groundwater contamination is believed to have arisen from a leaky pipe which carried aviation gasoline and jet fuel. Installation of piezometer nests, with short intake screens at multiple levels allowed estimation of high-resolution, vertical profiles of groundwater age using 3H/3He dating (Figure 19). Comparison between age and contaminant concentration data allowed the evolution of the groundwater plume to be determined. Vertical profiles of groundwater age allow determination of the vertical flow velocity. The plume defines a groundwater flow line, and comparison of ages within the center of the plume at different locations enables estimation of the horizontal flow velocity. With this information, the authors were able to trace the flow lines upgradient of the sampling points and identify both the location of the contaminant source and the approximate year that the contamination first occurred (Figure 43).

Figure showing groundwater flow lines derived from an analytical model
Figure 43 – Groundwater flow lines (blue dashed lines) derived from an analytical model calibrated to contaminant concentrations and 3H/ 3He age determinations (black squares denote sampling locations) at a contaminated site at Cape Cod, Massachusetts, USA. Given the groundwater age profile and the depth of the contaminant plume at three piezometer nests (ITW-1, ITW-2 and ITW-3), the authors were able to calculate the approximate location of the contaminant source and the year that the contamination occurred. The shaded region shows the flow lines originating at the contaminant location, and the red circles show the simulated movement of groundwater recharged at this location in 1975, when the contamination event was estimated to have occurred. Thus, the contaminant moved approximately 100 m in the first ten years (by 1985) and a further 400 m in the next ten years (by 1995). The vertical 3H and 3He profile obtained in piezometer nest ITW-3 is depicted in Figure 19 (After Solomon et al., 1995).