| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Bowdoin College, Brunswick, ME
2 Woods Hole Oceanographic Institution, Woods Hole, MA
* Corresponding author: mcharette{at}whoi.edu
Submarine groundwater discharge (SGD) is the flux of fresh and brackish groundwater to the ocean through a coastal aquifer (1). Groundwater often transports large amounts of anthropogenic nitrogen and phosphorus to the ocean, and understanding the magnitude of groundwater flux is important to the environmental protection and management of coastal waters (2, 3). Recent studies have employed radium isotopes (4) and radon-222 (5), both naturally enriched in groundwater relative to surface water due to a high uranium content in aquifer sediments, to quantify SGD in coastal systems. However, these studies have been limited by a poor understanding of the distribution of these isotopes in coastal groundwater, the processes controlling these distributions, and a lack of high-resolution time-series sampling of tracer activities in surface waters. Here, we attempt to overcome these earlier limitations by mapping the distribution of 222Rn (t1/2 = 3.83 days) and 226Ra (t1/2 = 1600 years) across a groundwater salinity gradient, and by deploying a new in situ 222Rn analyzer to study the time dependence of SGD in Waquoit Bay.
Using a drive-point piezometer, four depth profiles of groundwater 226Ra and 222Rn were collected along a transect perpendicular to the shore at the head of Waquoit Bay, creating a vertical cross section of each isotope at the groundwater-seawater interface (Fig. 1). At each sampling depth, 226Ra was extracted from 
15 l of groundwater on a column of Mn-impregnated fiber. In the laboratory, the fiber was ashed and 226Ra quantified via gamma ray spectroscopy (6). For 222Rn, 250 ml of groundwater was collected in a glass bottle and directly analyzed on a Durridge RAD7 electronic radon monitor; activities were decay-corrected to the time of collection. To quantify tracer concentrations in surface water, discrete bay water samples for 226Ra were collected at four locations at the head of the bay. Continuous measurements of 222Rn at a single station at the head of Waquoit Bay were obtained using the RAD7 coupled to an air-water equilibrator as described in Burnett and Dulaiova (5). The instrument recorded the 222Rn activity of surface water every 30 min for 3 days. Tidal height, measured as water column depth to the sea floor at the sampling location, was monitored with a YSI 600 series CTD, and local atmospheric conditions were obtained from a nearby NOAA weather tower (# BUZM3, Buzzards Bay, MA).
|
|
3-fold increases in activity at salinities higher than 20 (Fig. 2a). 222Rn also displayed an increase in activity with increasing depth and distance from the berm (Fig. 2b). Because Rn is a noble gas, its distribution throughout the aquifer should be uniform, and not affected by changes in salinity. Also, because the residence time of fresh groundwater beneath the bay is several years (8), the activities of 222Rn in this zone are likely to be in equilibrium with the sediment activity of its parent isotope 226Ra. Therefore, the higher observed activities of 222Rn in the saline portion of the aquifer must be a function of a greater 226Ra source, which is consistent with the observed 226Ra distribution.
To quantify SGD to Waquoit Bay, we applied a non-steady-state mass balance model to our time-series 222Rn record collected over 3 days at the head of the bay. Corrections for atmospheric evasion, loss due to advection to the lower bay, and changes in inventory due to water column depth were all applied to the time-series data using the approach of Burnett and Dulaiova (5) to calculate the excess 222Rn flux (dpm m-2 d-1) during each 30-min interval. Dividing by the average groundwater 222Rn activity (dpm m-3), we estimated SGD rates ranging from 0 to 0.16 m d-1, with a mean of 0.08 ± 0.02 m d-1 (n = 132, 1-
). In general, SGD was related to tidal stage (Fig. 2c), whereby SGD was highest at low tide and lowest at high tide, a result consistent with other time-series estimates of SGD in coastal systems (6, 9). We suggest that both a change in hydraulic head gradient, modulated by the rise and fall of the tide, and the effects of subsurface recirculation due to tidal pumping are the likely explanation for this observation.
To calculate a volumetric flux of SGD (m3 d-1), we must first estimate the area of the seepage face at the head of Waquoit Bay. The length of the seepage face (1760 m) was obtained from a false-color aerial infrared image (1-m resolution) taken at low tide in the fall of 2002 (Charette, unpubl. data). A seepage meter study by Michael et al. (10) found the width of the seepage face to be 70 m. Based on this effective seepage surface area of 123,200 m2, we estimated a volume discharge of 9900 m3 d-1. Following the mass balance model of Charette et al. (4) and using the same effective seepage area, we estimated a 226Ra-derived volumetric flux of SGD of 1200 m3 d-1 (n = 4). Since high levels of 226Ra are present only in the saline portion of the aquifer and 222Rn is high in both fresh and saline groundwater (relative to surface waters), it is possible that the difference between these two estimates represents the fresh component of SGD. However, an alternative explanation is greater uncertainty in the 226Ra estimate due to the limited data compared to the 222Rn record.
Recent applications of 222Rn and 226Ra to estimate SGD have assumed that groundwater discharge is static over time and space. The flux estimate derived from the continuous 222Rn record, however, suggests substantial temporal variability in groundwater discharge across a tidal cycle. This method would be useful for longer-term studies, where variability in SGD may be driven by the spring/neap tidal cycle and seasonal to interannual changes in aquifer recharge. Also, high-resolution sampling (cm scale) of 222Rn and 226Ra in groundwater revealed that these isotopes do not behave as predicted across a salinity gradient, probably because of an increased source term across the freshwater-saltwater interface.
We thank the Waquoit Bay National Estuarine Research Reserve (WBNERR) for the use of their field site and research facilities during the summer of 2003. This study was funded by a NSF-Research Experience for Undergraduates (REU) site grant (OCE-0097498) and a NSF grant (OCE-0095384) to M.A.C.
Literature Cited
This article has been cited by other articles:
|
|
 |
|
  J. M. Talbot, K. D. Kroeger, A. Rago, M. C. Allen, and M. A. Charette Nitrogen Flux and Speciation Through the Subterranean Estuary of Waquoit Bay, Massachusetts Biol. Bull., October 1, 2003; 205(2): 244 - 245. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
= 226Ra, O = 222Rn, * - ** = transect of 4 depth profiles). A vertical cross section of groundwater revealed a sharp increase in salinity at the freshwater-saltwater interface (C, contour plot from * to **).
)) reveals increasing activity from the surface to 8 m, and distance from the berm to 12 m. A 5-point running average of the time-series record of groundwater velocities is plotted alongside tidal height above sea floor (C).