Spatial variability of submarine groundwater discharge, Ubatuba, Brazil

Abstract

Submarine groundwater discharge (SGD) is delivered into the oceans at various hydrogeological settings. Where coastal aquifers have spatial heterogeneous properties such as preferential flowpaths, SGD is also commonly heterogeneously distributed along the coast. SGD from a heterogeneous, fractured rock aquifer on the Ubatuba coast, Brazil has been investigated by depth profiling of ground conductivity. On spatial scales of <10 m, a significant variation in sub-surface conductivity and associated discharge rate is documented. These geophysical observations were used to identify preferential flowpaths, and to explain significant differences in average discharge rates observed in seepage meter records placed only a few metres apart from each other. This paper illustrates the utility of sub-surface conductivity measurements for accurate investigations of SGD fluxes.

Introduction

Rivers are obvious and visible pathways for terrestrial runoff. In contrast, submarine groundwater discharge (SGD) is a poorly understood process hidden from view. Moreover, SGD as a source of freshwater and dissolved constituents is the least studied element of water, salt and nutrient budgets of the coastal oceans (e.g. Zektser, 2000). According to Johannes (1980), “SGD should occur anywhere that an aquifer is hydraulically connected with the sea through permeable rocks or bottom sediments and where the [density-corrected] head is above sea level”. These conditions are met in many, if not most coastal areas. Surprisingly, SGD has only recently been a focus of research, particularly where human activities at the coast might influence sensitive marine environments. Within the coastal plain and seafloor sediments, a complex mixing zone of freshwater and seawater exists. This zone has been termed a “subterranean estuary” (Moore, 1999), in recognition of the similarities of the physical and chemical processes occurring in such interfaces and surface estuaries.

A prime reason for the lack of scientific understanding of SGD is the logistical complications of measuring a subterranean flux of water at the coastal interface. Parallel to hydrological modelling, various experimental methods have been developed in recent years. Direct flux measurements are commonly obtained with benthic chambers or seepage meters deployed on the seafloor, based on a design by Lee (1977). In such studies, total fluxes are inferred by multiplying direct point measurements of flux per unit area with the total expected discharge area. Indirect water-borne tracer approaches estimate groundwater fluxes integrated along a coastline, whereby a mass balance for a tracer is closed (e.g. Burnett et al., 2002). Commonly used natural tracers are Radon, Radium, nutrients and trace elements, and occasionally, injected tracers such as SF₆ are used (Clark and Stieglitz, in press). Where flow paths are simple and flow patterns homogenous, directly and indirectly measured total flow rates compare well with modelled total flow (Burnett et al., 2002). However, where aquifer confinement or preferential flow paths exist, care has to be taken with inferring total fluxes from direct point measurements of SGD fluxes. In such cases, spatial variability of SGD could be significant, and therefore, a more detailed understanding of the spatial distribution of fresh and saltwater fluxes across the sediment-water interface is required to more accurately estimate total flow rates. Here we demonstrate the utility of a simple geoelectric method to identify spatial heterogeneity and preferential flow paths.

As part of an initiative developed by the International Atomic Energy Agency (IAEA) and UNESCO to assess methodologies and importance of SGD for coastal zone management, multidisciplinary studies of SGD were carried out in various hydrogeological settings, i.e. coastal plain aquifers in Western Australia, glacial aquifers on Long Island (USA), karstic aquifers in Sicily, fractured rock aquifers in Ubatuba, Brazil and lava tube aquifers on the island of Mauritius (e.g. Burnett et al., 2006). In order to characterise the spatial distribution of SGD at these study sites, the electrical conductivity of groundwater was mapped. Here we report geophysical investigations of the bulk ground conductivity at Ubatuba, Brazil carried out in November 2003, and discuss the utility of such quantitative investigations to improving flux estimates of SGD by conventional methods, e.g. with seepage meters.