Submarine groundwater discharge in coastal Mediterranean areas by using radium isotopes: The Venice Lagoon, Minorca and Castelló.

Student thesis: Doctoral thesis


This Thesis represents an effort to bring together land-based and marine hydrogeological research by focusing on a process that occurs on the edge of both disciplines: the Submarine Groundwater Discharge (SGD) in the coastal zone. Water and chemical fluxes from land to the sea have been reported to be of major interest for many years, although they have been typically focused on surficial riverine water fluxes. It was not until the mid-1990s that SGD drew the attention of scientists, and it has been of increasing consideration since then, as evidenced by the number of publications. SGD fluxes, that include fresh groundwater and recirculated seawater, have been demonstrated to be significantly important in several coastal settings around the world (e.g., Brazil, California, Florida, Hawaii, Korea, Mexico, Sicily, South Carolina, Thailand, etc) because they can represent an important source of freshwater and may contain elevated nutrient concentrations or high levels of potentially harmful contaminants, such as metals (Moore et al., 2008). However, submarine groundwater discharge in the coastal zone is, as yet, insufficiently studied and difficult to measure, thus justifying the present work.
One of the regions where advances in the understanding of SGD processes are necessary is the Mediterranean shoreline. In this area, coastal management is complex due to the continuous growth of coastal population and tourism that exploits and pollutes groundwater resources. Indeed, groundwater contamination and saltwater intrusion are problems that have already been recognized in several areas. The Mediterranean shoreline is also characterized by an abundance of karstic areas, which are potentially subject to significant SGD fluxes given their enhanced permeability. Information regarding this neglected aspect of coastal management is required for better planning, implementation of coastal policies and comprehension of the processes playing a role in the ecology of litoral regions.
Due to the difficulty of detecting and quantifying SGD, several techniques have been developed. One of the most successfully applied is based on the use of the four naturally occurring radium isotopes (223Ra, 224Ra 226Ra and 228Ra). Given their different half-lives, radium isotopes are suitable to trace SGD processes at several time scales. Consequently, radium has undergone a renaissance in ocean science studies. Concurrent with this new application was the production of a commercially available delayed coincidence counter (RaDeCC; Moore and Arnold, 1996), which improved and simplified the measurement of the two short-lived Ra isotopes, 224Ra and 223Ra.
Overall, the potential relevance of SGD in coastal regions, the likely ecological significance of SGD in many areas, the emergence of the Ra quartet as a promising tool to estimate this process together with the lack of studies on SGD fluxes around the Mediterranean basin have motivated the present work.
The overall goal of this dissertation is to employ radium isotopes to evaluate the SGD in different coastal settings in order to contribute to the understanding of Mediterranean coastal hydrology through tracer systematics. This general objective is addressed from three main viewpoints: i) technical examination, ii) methodological application and iii) implications assessment. Thereby, the specific aims are to:

i) examine the measurement technique for the short-lived radium isotopes, the Radium Delayed Coincidence counting system (hereafter, RaDeCC) and determine the associated uncertainties on an operation-by-operation propagation basis.

ii) test the utility of naturally occurring radium isotopic tracers to measure SGD by using different mass balance models and assess the absolute magnitude of the SGD water flux in three contrasting Mediterranean coastal areas: the northern Venice Lagoon, a karstic cove in Minorca Island and the Badum coastal area in Peníscola (Castelló).

iii) estimate the SGD fluxes of nutrients for the first time in a Mediterranean coastal region (Minorca and Badum sites) and try to decipher whether they are related to ecological aspects such as the occurrence of summer algal blooms in Minorca.

The main results have been classified in 1) Technical remarks and 2) Outcomes from the SGD studies and are the following:

1) Technical remarks
A relevant methodological contribution of this Thesis is the examination of the Radium Delayed Coincidence Counter (RaDeCC) system to measure 223Ra and 224Ra activities in water samples. Specifically, a detailed assessment of the propagated uncertainties throughout the complete process (from sampling to the final determination of radium activities) allowed concluding with recommendations for every user of the system to assess and reduce the final uncertainty of the measurements.
Asymptotic uncertainty levels are attained at ca. 7% for 223Ra and 4% for 224Ra, which could be obtained for typically radium-enriched groundwater samples. On the other hand, because of their low radium content, higher uncertainties could be acceptable for seawater samples. These results should serve as guidance for water sampling and radium measurements. For example, we showed that for the above listed uncertainties for waters with 223Ra ranging between 4 and 5 dpm/100 L and 224Ra from 20 to 40 dpm/100 L, a minimum volume of 35 L would be required, the 224Ra could be counted as late as 4 days after collection and the measurements should be longer than an hour. For the same sample, the second count for 223Ra should be performed no later than 16 days from sampling and the counting should last for at least 6 h.

2) Outcomes from the SGD studies
The contrasting Mediterranean coastal areas of interest were: an enclosed, tidal-influenced, water body (northern Venice Lagoon), a small and narrow cove (the Alcalfar Cove in Minorca) and an open karstic site (Badum, in Castelló). In all these regions SGD has been found to be the major source of radium to the coastal area. Different Ra-based approaches have been successfully tested and the absolute magnitudes of SGD have been determined:

i) in the northern Venice Lagoon, the estimated SGD rate in May-June 2004 averaged 3000000 m3/d, and it resulted to be dominated by recirculated sweater.
ii) in the Alcalfar Cove (Minorca), the total SGD in May 2005 and February 2006 was calculated to be 411±89 m3/d, and distributed in 123±36 m3/d of recirculated seawater and 288±82 m3/d of fresh groundwater.
iii) radium isotopes were applied together with dissolved inorganic silica and major elements in seawater to determine the total SGD flux to the Badum area for two different periods: 190000 m3/d for June 2007 and 40000 m3/d in October in 2006.
Northern Venice Lagoon
This study implied the evaluation of several sources of radium into the lagoon: riverine input, diffusion from fine grained sediments, coastal Adriatic Sea and SGD. Even though this complexity, the total SGD flux could be derived from the use of potential groundwater endmembers and for all the isotopes. Good agreement among the several SGD estimates was achieved, ranging from 40 to 260 L/m2/d.
Given that the estimated flux for the northern Venice Lagoon consisted of nearly wholly recirculated seawater (high salinity of the groundwater endmember), a conclusion to be highlighted here is the importance of marine forcing factors to enhance SGD in flat and tidally controlled coastal areas. Other studies in the literature have already appraised the significance of tidal pumping as either a forcing mechanism or a modulating factor (Kim and Hwang, 2002; Taniguchi, 2002; Rapaglia, 2005; Bokuniewicz et al., 2008; Li et al., 2009).
The estimated SGD flux was found to be very similar to the overall riverine input, therefore emphasizing the potential significance of SGD-associated chemical load to a sensitive ecosystem. Thus, a further recommendation is that potential contamination of discharging groundwater should engage accurate attention of water management activities.

Alcalfar Cove, Minorca
Results from the different surveys carried out in the Alcalfar Cove showed a clear water column stratification with a fresher buoyant layer of 0.3 m depth, affected by the SGD from coastal brackish springs. In contrast to the study in the Venice Lagoon, here only two sources affect the radium distribution in the cove: the brackish coastal springs and the coastal sea. After estimating a very minor contribution of radium from sands, a radium binary mixing model was applied. Furthermore, we were able to separate the total SGD in the two components of fresh (17 L/m2/d) and recirculated (7 L/m2/d) fractions.
An added value of this study was the determination of the SGD-associated input of new nutrients to the cove. They were estimated to be on the order of 18000 micromol/m2/d of NO3, 1140 micromol/m2/d of SiO4 and 4 micromol/m2/d of PO4. The significantly elevated fluxes of dissolved inorganic nitrogen could be due to human activities in the vicinity, particularly from agricultural and farming practices bringing N from fertilizers and manure to groundwater. It is important to note that only inorganic nutrients were assessed and, consequently, estimated fluxes were minimal values of the total nutrient SGD load, leaving aside the potential input of organic nutrients.
Since no riverine inputs are present in the study area, an additional concluding remark of this study resided on the likely connection between the nutrient inputs via SGD and the reported proliferation of dinoflagellate algal blooms in summer (>106 cells/L). The control of coastal waters quality in such setting is also important because of tourism uses during summer and, therefore, the role of SGD fluxes in Minorca coastal areas should receive more consideration.

Badum coastal area, Castelló
In the Badum area, the SGD study benefited from background such as a thermal infrared image obtained in June 2006 covering the whole are where the groundwater from the Maestrat Jurassic Aquifer is expected to discharge. The registered fresher plume is originated from coastal springs probably related to tectonic fractures. The SGD flux calculated in June 2007 (190000 m3/d or 86 L/m2/d) and October 2006 (40000 m3/d or 53 L/m2/d) allowed suggesting a time lag of approximately 3 months from precipitation in the recharge region to be manifested in the coastal zone as SGD. This interval indicated the mean residence time of groundwater inside the aquifer, which is actually lower (i.e., rapid discharge) than time-gaps determined by other researchers in non karstic aquifers (of up to half a year) not disrupted with preferential flow paths. As occurred in the Alcalfar Cove site, the fluxes of nutrients driven by SGD to coastal waters carried an elevated nitrogen load compared to the phosphorus signal (1500/19 and 8100/39for the autumn and summer periods, respectively), that could trigger significant changes in the nutrient equilibrium if the coastal waters and bring this environment to enhanced P limitation for primary production. Differently from the previous study, however, this input of nutrients is considered to be of natural origin, since the area is located inside a Natural Park, a declared natural protected area where most anthropogenic land uses are not permitted.

Finally, the main conclusions derived from the presented studies in Mediterranean areas could be summarized as follows:

- Radium isotopes are useful tracers of SGD in a wide range of hydrogeological settings.
- Total SGD is estimated with this proxie, although allowing one to distinguish between fresh groundwater and recirculated seawater fractions if the groundwater contributor can be sampled at both fresh and salty salinities.
- Water residence times can be successfully determined by short-lived radium distribution in the study area.
- Dissolved inorganic nutrient input via SGD may be important in several Mediterranean coastal areas, either anthropogenically altered or pristine, in terms of altering the biological equilibrium (inducing algal blooms or phosphorus limitation).
- Karstic areas present a groundwater residence time of approximately 3 months (i.e., time lag from rainfall to actual discharge), shorter than other homogeneous aquifers.
Date of Award15 Sept 2009
Original languageEnglish
SupervisorPere Masque Barri (Director) & Jordi Garcia Orellana (Director)

Cite this