Tritium, Helium, and Neon Data, and Apparent Ages

Tritium, helium, and neon data, and tritium/helium-3 apparent ages are presented in table 1. Apparent ages are reasonable for 10 of the 12 sites; apparent ages range from 1.1 to 21.2 years. Data for the other two wells indicate site-specific problems, and can not be reliably dated (see sections, “Helium-4 and Neon Data” and “Tritium Data”). Apparent ages derived from data collected in 1997 and 1998 agree well; absolute differences range from 0.2 to 5.8 years, with a median absolute difference of 1.2 years. Reasonable method precision also is demonstrated in replicate data (table 1).

Helium-4 and Neon Data

Helium-4 and neon concentrations are plotted in figure 2. For the most part, the various sources of helium-4 and neon are clearly seen in figure 2. A component from air-water solubility at 8°C is represented by the 8°C point on the air-water solubility curve. Additional helium and neon are derived from excess air. The component of excess air varies in each sample and is represented by movement along the excess air curve away from the air-water solubility curve. The radiogenic component of helium-4 also varies in each sample; it is represented by the departure away from (and to the right of) the excess air curve, and parallel to the abscissa (x-axis).

Data for well 01N/01E-07AAC (henceforth, 7AAC) is a low-yield well. Well 7AAC was sampled at a flow rate of 1.8 L/min (the well is unable to sustain a faster flow rate), resulting in a 4-minute time-of-travel through the 56 m of collection lines. Gas bubbles were observed in the collection lines, apparently a result of degassing during the long travel time in the collection lines. Degassing also might have occurred in the well if the water level in the well casing had been drawn down below the top of the screen, resulting in cascading water in the well; this is a possibility, given the fact that the well was repeatedly pumped dry during well development. Data from samples that have degassed are not suitable for tritium/helium-3 dating. Such data commonly plot above the excess air curve and tend to yield unreasonable or negative apparent ages. The 1997 sample lies above the excess air curve (fig. 2), and both the 1997 and 1998 samples have negative apparent ages (table 1). Meaningful apparent ages can not be assigned from the data for well 7AAC; therefore, the data are not interpreted further.

Tritium Data

Concentrations of tritiogenic helium-3 can be combined with measured tritium concentrations to calculate reconstructed (beginning-of-the-flowpath) tritium concentrations. Reconstructed tritium concentrations are compared to apparent recharge date (sampling date minus apparent age) in figure 3. Annual precipitation-weighted tritium concentrations of Portland precipitation for 1963–93 (International Atomic Energy Agency/World Meteorological Organization, 1998) also are shown in figure 3.

Most of the reconstructed tritium concentrations plot close to, but to the right of, the precipitation data (fig. 3). This offset commonly is observed and has been attributed to the time-of-travel through the unsaturated zone (Solomon and others, 1995; Cook and Solomon, 1997; Solomon and Cook, 2000). An apparent age is an estimate of the time of travel to a measurement point from the water table, not from land surface. Water just arriving at the top of the water table will have entered the unsaturated zone at some time in the past. The time of travel through the unsaturated zone can be on the order of years, and a response in water-table altitude during the rainy season does not indicate that the water being added to the water table from the bottom of the unsaturated zone is the same water that is entering the top of the unsaturated zone.

Another process that can cause apparent offset from the tritium input function for an aquifer is dispersion of tritium in the unsaturated zone. The tritium input function shown in figure 3 represents tritium in Portland precipitation; however, tritium crossing the water table will be affected by dispersion in the unsaturated zone (on the order of 24 m near the sampled wells). Dispersion of tritium results in moderation of the 1963 peak and an increase in later-time tritium concentrations relative to piston flow (for example, Solomon and Cook, 2000). Thus, at least some of the apparent offset from the tritium input function probably reflects actual dispersion in the unsaturated zone, which is not represented in figure 3.

Offset to the right of the precipitation data also is demonstrated (fig. 3) by data from well 01S/01E-13BDB (henceforth, 13BDB), a domestic well in the Portland basin sampled in 1993 for tritium (but not helium) as part of NAWQA activities (Hinkle, 1997). Well 13BDB is given consideration because it was screened near the top of the water table (from 0.0 to 0.8 m below the water table) when sampled. Thus, water from well 13BDB should have an age of essentially 0 years, and the reconstructed tritium concentration for this water sample should be essentially equal to the measured tritium concentration. Data for well 13BDB are offset from the precipitation data by about the same amount as the other reconstructed tritium data (except for well 01S/01E-02DBC), suggesting that these offsets are real, and not artifacts of the tritium/helium-3 method.

Data for one well, 01S/01E-02DBC (henceforth, 2DBC), plot below the tritium input function. Analytical precision is not the source of the problem with well 2DBC, as the data demonstrate reasonable precision for the six samples analyzed for this site (table 1). Offset below a tritium input function can be caused by mixing along flowpaths or in wells (Aeschbach-Hertig and others, 1998). Thus, data for well 2DBC probably represent mixing of water of different ages. For example, a mixture of 85 percent water recharged in 1947 (initial tritium concentration of 4 TU) with 15 percent water recharged in 1980 (initial tritium concentration of 35 TU) would yield the tritium and tritiogenic helium-3 concentrations observed in the 1997 and 1998 samples. This example is given only to demonstrate the effect of mixing on tritium/helium-3 apparent ages, and is not meant to imply that this mixing example represents the mixing proportions that actually occurred.

A mixing hypothesis for well 2DBC is supported by the observation that of the 12 wells sampled for this project, well 2DBC had the greatest penetration of the water table, likely increasing the opportunities for mixing in the aquifer. Water from well 2DBC probably is a mixture of water older and younger than the apparent recharge age. The apparent recharge age is considered unreliable. Water from this well is considered undatable by the tritium/helium-3 method, and the data are not further interpreted.

Age/Depth Relations

The relation between apparent age and depth (specifically, depth of the center of the well screen below the water table, or water table penetration) is presented in figure 4. The overall pattern is one of increasing apparent age with increasing depth.

Lines representing age/depth relations based on assumptions of effects of recharge rate on vertical ground-water movement also are shown in figure 4. The recharge-based age/depth relations are derived using the approach of Cook and Böhlke (2000). Briefly, for water close (relative to the thickness of the aquifer) to the water table in a recharge area of an unconfined aquifer, the time of travel can be approximated by:

, (1)

where

t is time of travel (year),

z is depth below water table (meters),

θ is porosity (unitless), and

R is recharge rate (meters per year).

Recharge-based estimates of age/depth relations were made using recharge rate estimates from a Portland basin recharge model (Snyder and others, 1994) and porosity estimates assembled for particle-tracking analysis in the Portland basin (Hinkle and Snyder, 1997; Snyder and others, 1998). Mean recharge rate and porosity estimates for the 10 model grid cells containing these wells were used in these calculations.

The Portland basin recharge model was a simulation of 1987–88 recharge conditions. Recharge in many core urban areas of the Portland basin likely has increased in the time since the late 1980s, in response to a program to divert more stormwater into the ground-water system using so-called drywells. To provide a measure of the possible effect of increased drywell recharge, an alternative recharge-based estimate of age/depth relations was made using recharge estimates revised to account for increased drywell density. This estimate was calculated using the methods of Snyder and others (1994) with the additional assumption that all impervious areas drain to drywells. Thus, this estimate represents the maximum recharge that would be simulated in the recharge model as a result of increased drywell density, and represents an upper bound to the estimate of recharge. Maximum-recharge-based age/depth relations also are shown in figure 4.

Age/depth relations based on original recharge estimates provide a reasonable fit to the tritium/helium-3 data (fig. 4), especially given the assumptions and uncertainties inherent in recharge-based age/depth relations. Ground-water samples with tritium/helium-3 apparent ages of less than 5 years demonstrate a somewhat closer fit to maximum-recharge-based age/depth relations. Although younger water would be most affected by recent changes in recharge rates, the apparent agreement between young (less than 5 year) tritium/helium-3 apparent ages and age/depth relations based on maximum-recharge estimates could be coincidental. Overall, the data shown in figure 4 support the original recharge estimates, but also suggest that effects of increased recharge from recently installed drywells might be starting to appear.