Earth-Science Reviews 220 (2021) 103681 Contents lists available at ScienceDirect Earth-Science Reviews journal homepage: www.elsevier.com/locate/earscirev Invited Review Radium isotopes as submarine groundwater discharge (SGD) tracers: Review and recommendations J. Garcia-Orellana a, b, *, V. Rodellas a, *, J. Tamborski c, M. Diego-Feliu a, P. van Beek d, Y. Weinstein e, M. Charette f, A. Alorda-Kleinglass a, H.A. Michael g, h, T. Stieglitz i, J. Scholten j a ICTA-UAB, Institut de Ciència i Tecnologia Ambientals, Universitat Autònoma de Barcelona, E-08193 Bellaterra, Spain b Departament de Física, Universitat Autònoma de Barcelona, E-08193 Bellaterra, Spain c Department of Ocean and Earth Sciences, Old Dominion University, Norfolk, VA, USA d Laboratoire d’Etudes en Géophysique et Océanographie Spatiales (LEGOS), Université de Toulouse, CNES/CNRS/IRD/Université Toulouse III Paul Sabatier, Toulouse, France e Department of Geography and Environmnet, Bar-Ilan University, Ramat-Gan, 52900, Israel f Department of Marine Chemistry and Geochemistry, Woods Hole Oceanographic Institution, Woods Hole, MA, USA g Department of Earth Sciences, University of Delaware, Newark, DE, USA h Department of Civil and Environmental Engineering, University of Delaware, Newark, DE, USA i Aix-Marseille Université, CNRS, IRD, INRAE, Coll France, CEREGE, Aix-en-Provence, France j Institute of Geosciences, Kiel University, Germany A B S T R A C T Submarine groundwater discharge (SGD) is now recognized as an important process of the hydrological cycle worldwide and plays a major role as a conveyor of dissolved compounds to the ocean. Naturally occurring radium isotopes (223Ra, 224Ra, 226Ra and 228Ra) are widely employed geochemical tracers in marine envi­ ronments. Whilst Ra isotopes were initially predominantly applied to study open ocean processes and fluxes across the continental margins, their most common application in the marine environment has undoubtedly become the identification and quantification of SGD. This review focuses on the application of Ra isotopes as tracers of SGD and associated inputs of water and solutes to the coastal ocean. In addition, we review i) the processes controlling Ra enrichment and depletion in coastal groundwater and seawater; ii) the systematics applied to estimate SGD using Ra isotopes and iii) we summarize additional applications of Ra isotopes in groundwater and marine studies. We also provide some considerations that will help refine SGD estimates and identify the critical knowledge gaps and research needs related to the current use of Ra isotopes as SGD tracers. 1. Introduction that depends on the U and Th content and the half-life of each Ra isotope. Radium (Ra) is the element number 88 in the Periodic Table and it Radium isotopes are widely recognized as important geochemical belongs to Group IIA, the alkaline earth metals. There are four naturally tracers in marine environments, mainly because i) they behave conser­ occurring radium isotopes (224Ra, 223Ra, 228Ra and 226Ra), which are vatively in seawater (i.e., lack of significant chemical and biological continuously produced by the decay of their Th-isotope parents of the U additions or removals); ii) they decay at a known rate and iii) they are and Th decay series (228Th, 227Th, 232Th and 230Th, respectively). The primarily produced from water-rock and sediment-water interactions. most abundant isotopes are 226Ra from the 238U decay chain, an alpha- Consequently, Ra isotopes have been traditionally used to trace land- gamma emitter with a half-life of 1600 y, and 228Ra from the 232Th ocean interaction processes (e.g., Charette et al., 2016; Elsinger and decay chain, a beta emitter with a half-life of 5.8 y. There are two short- Moore, 1983a, 1983b; Knauss et al., 1978), which are also referred to as lived isotopes that also occur naturally, 223Ra and 224Ra, which are boundary exchange processes (e.g., Jeandel, 2016), as well as to esti­ alpha-gamma emitters from the 235U and 232Th decay chains with half- mate mixing rates, apparent ages or residence times, in particular in lives of 11.4 d and 3.6 d, respectively. Uranium and thorium are widely coastal and ocean waters (e.g., Annett et al., 2013; Burt et al., 2013a; Ku distributed in nature, mainly in soils, sediments and rocks, and thus the and Luo, 2008; Moore, 2000a, 2000b; Moore et al., 2006; Tomasky- four Ra isotopes are continuously produced in the environment at a rate Holmes et al., 2013), but also in aquifers (e.g., Diego-Feliu et al., 2021; * Corresponding author at: ICTA-UAB, Institut de Ciència i Tecnologia Ambientals, Universitat Autònoma de Barcelona, E-08193 Bellaterra, Spain. E-mail addresses: jordi.garcia@uab.cat (J. Garcia-Orellana), Valenti.Rodellas@uab.cat (V. Rodellas). https://doi.org/10.1016/j.earscirev.2021.103681 Received 29 December 2020; Received in revised form 23 April 2021; Accepted 11 May 2021 Available online 14 May 2021 0012-8252/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Kraemer, 2005; Liao et al., 2020; Molina-Porras et al., 2020) and hy­ (Simpson et al., 1911). However, it was not until the publication of the drothermal systems (e.g., Kadko and Moore, 1988; Neuholz et al., 2020). first profile of Ra activities in seawater and sediments off the coast of Whilst initially applied to study open ocean processes and to estimate California by Evans and Kip (1938), which suggested the influence of different exchange processes across the continental margins, the most sediments on the increase of seawater 226Ra activities with depth, that common application of Ra isotopes in the marine environment has un­ the use of Ra isotopes to study oceanographic processes was considered. doubtedly been the quantification of water and solute fluxes from land The first true application of Ra as an oceanic tracer was by Koczy (1958), to the coastal sea, driven by Submarine Groundwater Discharge (SGD) who showed that the primary oceanic source of 226Ra was 230Th decay in (Fig. 1) (Ma and Zhang, 2020). Radium isotopes have been used to marine sediments, and who initiated the use of one-dimensional vertical quantify SGD in a wide range of marine environments, including sandy diffusion models to describe the upward flux and characteristic con­ beaches (e.g., Bokuniewicz et al., 2015; Evans and Wilson, 2017; centration profiles of 226Ra. Shortly thereafter, Koczy (1963) estimated Rodellas et al., 2014), bays (e.g., Beck et al., 2008; Hwang et al., 2005; that approximately 1–5% of the actual 226Ra present in seawater was Lecher et al., 2016; Zhang et al., 2017), estuaries (e.g., Luek and Beck, supplied by rivers. This was followed by a study of water column mixing 2014; Wang and Du, 2016; Young et al., 2008), coastal lagoons (e.g., and ventilation rates in the different oceans (Broecker et al., 1967). A Gattacceca et al., 2011; Rapaglia et al., 2010; Tamborski et al., 2019), global-scale picture of the distribution of long-lived Ra isotopes (226Ra salt marshes (e.g., Charette et al., 2003), large contintental shelves (e.g., and 228Ra) was first provided by the Geochemical Ocean Sections Study Liu et al., 2014; Moore, 1996; Wang et al., 2014), ocean basins (e.g., (GEOSECS – 1972–1978; e.g., Chan et al., 1977; Trier et al., 1972). Moore et al., 2008; Rodellas et al., 2015a) and the global ocean (e.g., In the mid 1960s, Blanchard and Oakes (1965) reported elevated Kwon et al., 2014). SGD is now recognized as an important process activities of 226Ra in coastal waters relative to open ocean waters. The worldwide and plays a major role as a conveyor of dissolved compounds detection of significant activities of 228Ra in both surface and bottom to the ocean (e.g., nutrients, metals, pollutants) (Cho et al., 2018; Moore, waters led to the understanding that the distribution of Ra depended on 2010; Rodellas et al., 2015a). Thus, SGD has significant implications for sediments but also on other sources that introduced Ra into the oceans coastal biogeochemical cycles (e.g., Adolf et al., 2019; Andrisoa et al., (Kaufman et al., 1973; Moore, 1969a, 1969b). Few years later, Li et al. 2019; Garces et al., 2011; Garcia-Orellana et al., 2016; Ruiz-González (1977) observed higher concentrations of 226Ra in the Hudson River et al., 2021; Santos et al., 2021; Sugimoto et al., 2017). This review is estuary compared with those observed either in the river itself or in the focused on the application of Ra isotopes as tracers of SGD-derived in­ adjacent surface ocean water, proposing that 226Ra was released by puts of water and solutes to the coastal ocean but also aims to provide estuarine and continental shelf sediments, which represents an impor­ some considerations that will help refine SGD estimates and identify tant source of 226Ra to the ocean. This was further corroborated by critical knowledge gaps and research needs related to the current use of measurements in the Pee Dee River-Winyah Bay estuary, in South Car­ Ra isotopes as SGD tracers. olina, USA, by Elsinger and Moore (1980) who concluded that Ra desorption from sediments could quantitatively explain the increase of 226 2. Radium isotopes used as SGD tracers: historical perspective Ra in brackish water. Several studies followed these initial in­ vestigations, highlighting the importance of coastal sediments in the 2.1. Early days: radium isotopes as tracers of marine processes release of long-lived Ra isotopes to coastal waters (e.g., Li and Chan, (1900–1990) 1979; Moore, 1987). Measurement of 226Ra and 228Ra in seawater samples was relatively Radium was discovered from pitchblende ore in 1889 by Marie complex prior to the wide-scale availability of high-purity germanium Sklodowksa-Curie and Pierre Curie, representing one of the first ele­ (HPGe) gamma detectors and the extraction of Ra from seawater using ments discovered by means of its radioactive properties (Porcelli et al., manganese-impregnated acrylic fibers. While 226Ra was measured in a 2014). This historic discovery initiated a remarkable use of Ra for few liters of seawater (5–10 L) by the emanation of 222Rn (Mathieu et al., medical, industrial and scientific purposes, including its use as an ocean 1988), the measurement of 228Ra required greater volumes (often >> tracer. Early oceanic investigations revealed elevated 226Ra activities in 100 L). The quantification of 228Ra proceeded either via the extraction of deep-sea sediments (Joly, 1908), but the first measurement of 226Ra 228 Ac and the measurement of its beta activity or via the partial activities in seawater was performed in 1911 to characterize the pene­ ingrowth of 228Th over a period of at least 4–12 months and the mea­ trating radiation observed in the ionization chambers of ships at sea surement via the alpha recoil of 228Th or via its daughter 224Ra by a Fig. 1. Record of published research articles that used the keywords 1) ‘radium’ and ‘submarine groundwater discharge’ (n = 511) and 2) ‘radium’ and ‘ocean’ (n = 477), as indexed by Web of Science. Articles related to NORM, radiation protection or sediment concentrations have been removed from the record of ‘radium’ and ‘ocean’. 2 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 proportional counter (Moore, 1981; Moore, 1969a, 1969b). The intro­ previously demonstrated in a hydrology-based study in Great South Bay duction of manganese-impregnated acrylic fibers for sampling Ra iso­ (NY, USA) conducted by Bokuniewicz (1992), and it decisively topes in the ocean was also a key advance in the use of Ra as a tracer in contributed towards defining the term SGD. marine environments (Moore, 1976; Moore and Reid, 1973). Before the Awareness of the volumetric and chemical importance of SGD was development of this approach, large volumes of seawater and long and significantly increased by Moore (1996), who linked 226Ra enrichments laborious procedures were required to concentrate Ra isotopes via in coastal shelf waters of the South Atlantic Bight to large amounts of BaSO4 precipitation (Kaufman et al., 1973; Moore, 1969a, 1969b; direct groundwater discharge. Moore suggested that the volume of SGD Moore, 1972; Trier et al., 1972). Since then, Ra isotopes in marine wa­ over this several hundred-kilometer coastline was comparable to the ters are concentrated in situ with minimum time and effort by observed discharge from rivers, although the SGD probably included manganese-impregnated acrylic fibers. More recently, mass spectrom­ brackish and saline groundwater. In a commentary on this article, etry has been used for 226Ra determination in seawater, such as thermal Younger (1996) questioned the inclusion of salty groundwater as a ionization mass spectrometry (TIMS) (e.g., Ollivier et al., 2008), mul­ component of SGD, and its comparison to freshwater discharge to the ticollector inductively coupled plasma mass spectrometry (MC-ICP-MS) coastal zone via rivers. The rebuttal by Church (1996) argued that SGD, (e.g., Te Hsieh and Henderson, 2011) or single-collector sector field regardless of its salinity, is chemically distinct from seawater and (ICP-MS) (e.g., Vieira et al., 2021). therefore can exert a significant control on coastal ocean biogeochem­ The application of the short-lived 223Ra and 224Ra as tracers of ma­ ical cycles. rine processes lagged the long-lived Ra isotopes by several years due to In the same year, Moore and Arnold (1996) published the design of analytical limitations and because their potential use in oceanic studies the Radium Delayed Coincidence Counter (RaDeCC), an alpha detector was limited due to their shorter half-lives (11.4 d and 3.6 d for 223Ra and system that allowed a simple and reliable determination of the short- 224 Ra, respectively). Short-lived Ra isotopes were first applied to estu­ lived Ra isotopes (223Ra and 224Ra) based on an original design of Gif­ arine mixing studies, which have time-scales of 1–10 days. Elsinger and fin et al. (1963). Since its introduction, the RaDeCC system has become Moore (1983a, 1983b) published one of the first studies on the distri­ the reference technique for quantifying short-lived radium isotopes in bution of 224Ra, 226Ra and 228Ra, which was focused on the mixing zone water samples, as it allows a robust and rapid 223Ra and 224Ra deter­ of the Pee Dee River-Winyah Bay and Delaware Bay estuaries (USA). mination with a simple setup at a relatively low cost. The full potential They showed that the main source of Ra isotopes were desorption and of short-lived Ra isotopes was described shortly after, when Moore diffusion from suspended and bottom sediments, which together published two studies providing conceptual models to derive ages of contributed to the non-conservative increase of the three isotopes in the coastal waters and estimate coastal mixing rates using 223Ra and 224Ra river-sea mixing zone. Shortly thereafter, Bollinger and Moore (1984) (Moore, 2000a,b). suggested that 224Ra and 228Ra fluxes in a salt marsh from the US eastern In the period of 1990–2000, these pioneering researchers opened a coast could be supported by bioirrigation and bioturbation (i.e., sedi­ new era in Ra isotope application to SGD quantification, including the ment reworking and pumping driven by benthic fauna, respectively). development of most of the conceptual models, approaches and tech­ Levy and Moore (1985) concluded that the potential use of 224Ra as a niques that are currently in use. tracer required a much better understanding of its input functions in coastal zones. The authors classified Ra sources into two main groups: 2.3. Expansion period: the widespread application of Ra isotopes as SGD (1) primary sources that include desorption from estuaries and salt tracers marsh particles and; (2) secondary sources such as dissolved 228Th present in the water column, longshore currents that may transport The pioneering work on Ra isotopes of the 1990s, the increased 224 Ra from other areas and in situ production from 228Th decay adsorbed understanding of the importance of SGD in coastal biogeochemical cy­ on suspended particles or bottom sediments. cles, the technical improvements of Ra determination via HPGe and the commercialization of the first RaDeCC systems led to a rapid increase in 2.2. Development period: radium isotopes as SGD tracers (1990–2000) the number of studies using Ra isotopes as tracers of SGD (Fig. 1). Hundreds of scientific articles have been published since then. During Whilst the concept of SGD had been introduced in several early the 2000s, while some studies only used long-lived Ra isotopes as SGD hydrological-based studies (e.g., Bokuniewicz, 1992; Bokuniewicz, tracers (e.g., Charette and Buesseler, 2004; Kim et al., 2005; Krest et al., 1980; Capone and Bautista, 1985; Cooper, 1959; Freeze and Cherry, 2000), most studies combined measurements of short-lived Ra isotopes 1979; Glover, 1959; Johannes, 1980; Kohout, 1966; Toth, 1963), the to estimate water residence times (or mixing rates) and long-lived Ra role of groundwater as a major conveyor of Ra isotopes to the coastal isotopes to quantify SGD fluxes (e.g., Charette et al., 2003; Charette ocean was not established until the benchmark papers of Burnett et al. et al., 2001; Kelly and Moran, 2002; Moore, 2003). However, a few (1990), Veeh et al. (1995) and Rama and Moore (1996). Burnett et al. studies quantified SGD fluxes by using only short-lived Ra isotopes (1990) suggested that the high 226Ra activities in water from the (Fig. 2; Boehm et al., 2004; Krest and Harvey, 2003; Paytan et al., 2004). Suwannee estuary (USA) and offshore were most likely supplied by One of the first studies where SGD fluxes were derived exclusively from submarine springs or seeps. Veeh et al. (1995) demonstrated that the short-lived Ra isotopes (223Ra and 224Ra) was conducted at Huntington “traditional” Ra sources to coastal waters (e.g., rivers, sediments) were Beach (CA, USA) by Boehm et al. (2004). Their flux estimates based on insufficient to support the 226Ra activities in waters from the Spencer 223 Ra and 224Ra were supported by a later analysis of 226Ra at this Gulf (South Australia), suggesting an external source for the excess location and by the results of a hydrological numerical model (Boehm 226 Ra, such as the “submarine discharge of groundwater from granitic et al., 2006). At the same period, Hwang et al. (2005) and Paytan et al. basement rocks”. The importance of groundwater as a source of Ra to the (2006) used only short-lived Ra isotopes to estimate SGD fluxes of water coastal sea led Rama and Moore (1996) to apply for the first time the and associated nutrients. Over the last decade there has been a growing four naturally occurring Ra isotopes (which they defined as the radium volume of studies using only the short-lived Ra isotopes as tracers of SGD quartet) as tracers for quantifying groundwater flows and water ex­ (e.g., Baudron et al., 2015; Ferrarin et al., 2008; Garcia-Orellana et al., change (North Inlet salt marsh, South Carolina, USA). One key finding of 2010; Krall et al., 2017; Shellenbarger et al., 2006; Tamborski et al., this study was that extensive mixing between fresh groundwater and 2015; Trezzi et al., 2016). This relative abundance of publications, in saline marsh porewater occurred in the subsurface marsh sediment which only short-lived Ra isotopes are used, is in contrast with those before discharging to the coastal ocean, resulting in a groundwater conducted in previous decades, in which long-lived Ra isotopes were the discharge that was not entirely fresh but included a component of most applied tracers (Fig. 2). seawater circulating through the coastal aquifer. This concept was The methodological advances in the measurement of short-lived Ra 3 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Fig. 2. Number of research articles published in 5-year periods in which Ra isotopes have been used as tracers of SGD. The articles were classified according to the Ra isotopes used: only short-lived (223Ra and/or 224Ra), only long-lived (226Ra and/or 228Ra) or a combination of both. The search was performed using the words “SGD or Submarine Groundwater Discharge” AND “Ra or Radium” as keywords in “Web of Science” (n = 477). Reviews, method articles and studies not applying Ra isotopes as tracer of SGD-related processes were excluded from this classification. A total of 286 articles were considered in this classification analysis. isotopes have also contributed to their widespread application as tracers associated solutes across the sediment-water interface, which are of other coastal processes, such as water and solute transfer across the commonly referred to as porewater exchange (PEX) or benthic fluxes, sediment-water interface (also referred to as porewater exchange (Cai and are sometimes excluded from the SGD definition because of its short et al., 2014), groundwater age as well as flow rates in coastal aquifers length scale (<1 m) (e.g., Moore, 2010; Santos et al., 2012). We are also (Kiro et al., 2013; Kiro et al., 2012; Diego-Feliu et al., 2021), secondary implicitly considering that groundwater (defined as “any water in the permeability in animal burrows (Stieglitz et al., 2013), transfer of ground”) is synonymous with porewater (Burnett et al., 2003) and that sediment-derived material inputs (Burt et al., 2013b; Sanial et al., 2015; the term “coastal aquifer” includes permeable marine sediments. The Vieira et al., 2020; Vieira et al., 2019), and horizontal and vertical coastal term “subterranean estuary” (STE), which is widely used in the SGD water mixing rates (Annett et al., 2013; Charette et al., 2007; Colbert and literature, is also used in this review to refer to the part of the coastal Hammond, 2007). Numerous new uses of the RaDeCC system have aquifer that dynamically interacts with the ocean (Duque et al., 2020; occurred in recent years, including the development of approaches to Moore, 1999), determined in the hydrological literature as the fresh- measure 223Ra and 224Ra in sediments (Cai et al., 2012; Cai et al., 2014) or saline water interface or the mixing zone. 228 Ra and 226Ra activities in water (Diego-Feliu et al., 2020; Geibert et al., Within this broad definition of SGD we incorporate disparate water 2013; Peterson et al., 2009; Waska et al., 2008), as well as quantification flow processes, some involving the discharge of fresh groundwater and advancements and guidelines (Diego-Feliu et al., 2020; Moore, 2008; others encompassing the circulation of seawater through the subterra­ Selzer et al., 2021). There is still a widespread application of long-lived Ra nean estuary, or a mixture of both. These processes can be grouped into isotopes, particularly to quantify SGD occurring over long flow paths five different SGD pathways according to the characteristics of the (Moore, 1996; Rodellas et al., 2017; Tamborski et al., 2017a), SGD into processes (George et al., 2020; Michael et al., 2011; Robinson et al., entire ocean basins or the global ocean (Cho et al., 2018; Kwon et al., 2018; Santos et al., 2012) (Fig. 3): 1) Terrestrial groundwater 2014; Moore et al., 2008; Rodellas et al., 2015a) and as a shelf flux gauge discharge (usually fresh groundwater), driven by the hydraulic gradient (Charette et al., 2016). Some investigations have also considered how to between land and the sea; 2) Density-driven seawater circulation, combine short- and long-lived Ra isotopes to discriminate between caused by either density gradients along the freshwater-saltwater different SGD flow paths and sources (e.g., Charette et al., 2008; Moore, interface, or thermohaline gradients in permeable sediments; 3) Sea­ 2003; Rodellas et al., 2017; Tamborski et al., 2017a). sonal exchange of seawater, driven by the movement of the freshwater-saltwater interface due to temporal variations in aquifer 3. Submarine groundwater discharge: terminology recharge or sea level fluctuations; 4) Shoreface circulation of seawater, including intertidal circulation driven by tidal inundation (at Initially, the concept of groundwater discharge into the oceans had beach faces, salt marshes or mangroves) and wave set-up; and 5) cm- been investigated largely by hydrologists, who considered the discharge scale porewater exchange (PEX), driven by disparate mechanisms of meteoric groundwater as a component of the hydrological cycle such as current-bedform interactions, bioirrigation, tidal and wave (Manheim, 1967). Thereafter, with the involvement of oceanographers pumping, shear flow, ripple migration, etc. Notice that whereas all of interested in the influence of SGD on the chemistry of the ocean, the these processes force water flow through the sediment-water interface, definition evolved, with SGD now encompassing both meteoric ground­ the discharge of terrestrial groundwater (Pathway 1) and, to a lesser water and circulated seawater (Bokuniewicz, 1992; Church, 1996; extent, density-driven seawater circulation (Pathway 2), which also Moore, 1996; Rama and Moore, 1996). The interest in SGD has increas­ contains a fraction of freshwater, are the only mechanisms that repre­ ingly grown within the scientific community as it is now recognized as an sent a net source of water to the coastal ocean. Pathways 3–5 can be important pathway for the transport of chemical compounds between broadly classified as “saline SGD”; the summation of all five pathways is land and ocean, which can strongly affect marine biogeochemical cycles generally considered “total SGD”. at local, regional and global scales (e.g., Cho et al., 2018; Luijendijk et al., 2020; Rahman et al., 2019; Rodellas et al., 2015a; Santos et al., 2021). 4. Mechanisms controlling Ra in aquifers and SGD SGD is also recognized to provide a wide range of ecosystem goods and services (Erostate et al., 2020; Alorda-Kleinglass et al., 2021) One of the main characteristics that makes Ra isotopes useful tracers The definition of SGD has been discussed in several review papers in of SGD is that coastal groundwater is often greatly enriched with Ra the marine geosciences field (e.g., Burnett et al., 2002; Burnett et al., isotopes relative to coastal seawater (Burnett et al., 2001). The enrich­ 2001; Burnett et al., 2003; Church, 1996; Knee and Paytan, 2011; ment of groundwater with Ra isotopes is originated from the interaction Moore, 2010; Santos et al., 2012; Taniguchi et al., 2019; Taniguchi et al., of groundwater with rocks, soils or minerals that comprise the geolog­ 2002). Here, we will adopt the most inclusive definition, where SGD ical matrix of the coastal aquifer. Some studies have summarized the represents “the flow of water through continental and insular margins from various sources and sinks of Ra isotopes in coastal groundwater (Kiro the seabed to the coastal ocean, regardless of fluid composition or driving et al., 2012; Krishnaswami et al., 1982; Luo et al., 2018; Porcelli et al., force” (Burnett et al., 2003; Taniguchi et al., 2019). We are thus 2014; Tricca et al., 2001). The most common natural processes that including those centimeter-scale processess that can transport water and regulate the activity of Ra isotopes in groundwater are: 1) radioactive 4 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Fig. 3. Conceptual diagram of an unconfined coastal aquifer including the major Submarine Groundwater Discharge pathways, subdivided according to the driving mechanism: 1) Terrestrial groundwater discharge (usually fresh groundwater); 2) Density-driven seawater circulation; 3) Seasonal exchange of seawater; 4) Shoreface seawater circulation; and 5) cm-scale porewater exchange (PEX). Pathways 1, 2 and 3 could be extended farther offshore in systems with confined units. production from Th isotopes and decay; 2) adsorption and desorption fraction) and mainly the Ra produced in the effective alpha recoil zone from the aquifer solids and 3) weathering and precipitation. The transit (i.e., by the Th bound in the outer mineral lattice and by the surface- time of groundwater in the coastal aquifer can also regulate the Ra ac­ bound, i.e., adsorbed Th) (Fig. 4). The decay of Th in the effective tivity in groundwater (Fig. 4). alpha recoil zone mobilizes part of the generated Ra from this zone to the adjacent pore fluid due to the alpha-decay recoil energy (Sun and Semkow, 1998; Swarzenski, 2007; Fig. 4). The extent of the effective 4.1. Radioactive production and decay alpha recoil zone is specific to each type of solid (i.e., mineralogy) and is a function of the size and of the surface characteristics of the aquifer Ra activity in groundwater is controlled, in part, by the production solid grains (Beck and Cochran, 2013; Sun and Semkow, 1998; Swar­ and decay rates of each radionuclide in both the groundwater and the zenski, 2007; Diego-Feliu et al., 2021). The greater the specific surface geological matrix (Fig. 4). Radium isotope production depends on the area of solids in the aquifer, the greater the fraction of Th in the effective continuous decay of Th isotopes (230Th, 232Th, 228Th, and 227Th) from recoil zone and thus, the pool of Ra available for solid-solution exchange the U and Th decay chains, which are predominantly contained in (Copenhaver et al., 1993; Porcelli and Swarzenski, 2003). The influence aquifer solids. Therefore, the U and Th content of the geological matrix of alpha recoil can produce deviations in the groundwater Ra isotopic regulates the production rate. Concentrations of U and Th can vary ratios in relation to that expected from host rock ratios, since each Ra significantly depending on the host material (igneous rocks: 1.2–75 isotope is generated after a different number of decay events in each of Bq⋅kg− 1 of U and 0.8–134 Bq⋅kg− 1 of Th; metamorphic rocks: 25–60 the decay chains. For instance, when Ra isotopes in groundwaters are in Bq⋅kg− 1 of U and 20–110 Bq⋅kg− 1 of Th; and sedimentary rocks: equilibrium with aquifer solids, the alpha recoil process alone may <1.2–40 Bq⋅kg− 1 of U and < 4 and 40 Bq⋅kg− 1 of Th) (Ivanovich and produce equilibrium 226Ra/228Ra ratios up to 1.75 that of the host rock Harmon, 1992). For instance, carbonate minerals are enriched in U 238 U/232Th ratio and 224Ra/228Ra equilibrium ratios ranging from 1 to relative to Th, while clays are enriched in Th. As a consequence, 2.2 (e.g., Krishnaswami et al., 1982; Davidson and Dickson, 1986; groundwater flowing through formations with different lithologies have Swarzenski, 2007; Diego-Feliu et al., 2021). different activity ratios of Ra isotopes (e.g., 228Ra/226Ra AR, where AR Several methods are described in the literature determining how Ra means activity ratio), which can be used to identify and distinguish recoil can be estimated. These methods can be classified into five groups: groundwater inflowing from different hydrogeologic units (Charette and (1) experimental determination of emanation rates of daughter radio­ Buesseler, 2004; Moore, 2006a; Swarzenski et al., 2007a) (see Section nuclides from aquifer solids into groundwater (Hussain, 1995; Rama and 6). On the other hand, radioactive decay depends only on the activity of Moore, 1984); (2) experimental determination of the parent radionu­ Ra isotopes in groundwater and their specific decay constants. clide within the alpha-recoil zone (Cai et al., 2014; Cai et al., 2012; Not all of the Ra produced in the aquifer is directly transferred to Tamborski et al., 2019); (3) theoretical calculations focused on deter­ groundwater. The fraction of available Ra (i.e., the exchangeable Ra mining the alpha-recoil supply based on properties of the host material pool) includes the Th dissolved in groundwater (usually a negligible 5 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Fig. 4. Conceptual model of the main processes determining the abundance of Ra isotopes in groundwater from a coastal aquifer. Blue and brown colour represents the liquid and solid phases, respectively. The light brown color represents the effective alpha recoil zone for Ra isotopes, which depends on the aquifer solids and the energy of the Ra alpha particle (the recoil distance usually ranges, on average, between 300 and 400 Å (Sun and Semkow, 1998)). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) (e.g., density, surface area; Kigoshi, 1971; Semkow, 1990; Sun and groundwater table elevation and marine driving forces (e.g., tides, Semkow, 1998); (4) mathematical fitting to advective transport models waves and storms, mean sea-level). Other physico-chemical properties (Krest and Harvey, 2003), and (5) in-situ determination via supply rate of groundwater, such as temperature and pH may also control the solid- of 222Rn (Copenhaver et al., 1992; Krishnaswami et al., 1982; Porcelli solution partitioning (adsorption/desorption) of Ra in coastal aquifers. and Swarzenski, 2003). Beck and Cochran (2013) reviewed the role of temperature and pH, concluding that while there is a Ra adsorption with increasing pH, there is no clear effect of temperature on the adsorption of Ra onto aquifer 4.2. Adsorption and desorption solids. In order to estimate the relative distribution of Ra between solid and In fresh groundwater, Ra is usually and preferentially bound to solution, a distribution coefficient, KD [m3⋅kg− 1], is commonly used. aquifer solids, although a small fraction is found in solution as Ra2+. The This coefficient is defined as the ratio of the cocentration of Ra on the adsorption of Ra on solid surfaces depends on the cation exchange ca­ solid surface per mass of solid [Bq⋅kg− 1] to the amount of Ra remaining pacity (CEC) of the aquifer solids, but also on the chemical composition in mass of the solution at equilibrium [Bq⋅m− 3], and considers the Ra of groundwater. The higher the CEC, the higher the content of Ra that chemical equilibrium processes of adsortion-desorption (KD = Raad­ may be potentially adsorbed on the grain surface of aquifer solids (e.g., sorbed/Radesorbed), but not the processes of weathering. The distribution Beck and Cochran, 2013; Beneš et al., 1985; Kiro et al., 2012; Nathwani coefficient is one of the most relevant parameters for understanding the and Phillips, 1979; Vengosh et al., 2009). The ionic strength of the so­ Ra distribution in coastal aquifers and for applying transport models of lution, governed mainly by the groundwater salinity, has been recog­ Ra in groundwater. Radium distribution coefficient values span several nized as the most relevant factor controlling the exchange of Ra between orders of magnitude depending on the composition of groundwater and solid and groundwater (Beck and Cochran, 2013; Gonneea et al., 2013; aquifer solids, ranging widely from 10-2 to 102 m3⋅kg-1 (Kumar et al., Kiro et al., 2012; Webster et al., 1995). High ionic strength (i.e., high 2020; Beck and Cochran, 2013). Many authors analyzed the influence of salinity) hampers the adsorption of Ra2+ due to competition with other different solid and water compositions within different experimental cations dissolved in groundwater (e.g., Na+, K+, Ca2+) and promotes the settings. The distribution coefficient is commonly determined by batch desorption of surface-bound Ra due to cationic exchange. As a conse­ experiments (e.g., Beck and Cochran, 2013; Colbert and Hammond, quence, Ra is typically more enriched in brackish to saline groundwater 2008; Gonneea et al., 2008; Rama and Moore, 1996; Tachi et al., 2001; than in fresh groundwater. Exceptions may include carbonate aquifers Tamborski et al., 2019; Willett and Bond, 1995). The distribution co­ where mineral dissolution is more important than desorption in driving efficient of Ra has also been determined by other methods such as groundwater Ra activities. Thus, Ra activities in coastal groundwater adsorption-desorption modelling (e.g., Copenhaver et al., 1993; Webster may vary substantially, depending on the subsurface salinity distribu­ et al., 1995); chromatographic columns tests (e.g., Meier et al., 2015; tion, which is dynamic due to the interaction between the inland 6 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Relyea, 1982); and chemical equilibrium calculations (Puigdomenech et al., 2009). In SGD studies the groundwater transit time is instrumental and Bergstrom, 1995). to understand the degree of Ra isotope enrichment in groundwater since it entered the aquifer through one of its boundaries (e.g., aquifer 4.3. Weathering and precipitation recharge of freshwater or seawater infiltration through permeable sed­ iments). Groundwater transit time largely depends on the hydraulic Radium isotopes are also supplied to groundwater by weathering conductivity of the system and the physical mechanisms that are forcing processes, including dissolution and breakdown of rocks or minerals the groundwater advective flow. The hydraulic conductivity describes containing Ra that may occur during the flow of groundwater through the transmissive properties of a porous medium (for a given fluid) and the coastal aquifer. The intensity of weathering processes mainly de­ thus clearly influences the groundwater flow velocity (Freeze and pends on the physicochemical properties of groundwater, such as tem­ Cherry, 1979). For instance, hydraulic conductivity in unconsolidated perature, pH, redox potential or ionic strength, and on the silty-clay aquifers is typically on the order of 1 - 100 cm⋅d− 1, while in characteristics of the geological matrix (e.g., mineralogy, specific sur­ unconsolidated sandy aquifers it is significantly higher (102 - 104 cm⋅d-1; face area of aquifer solids) (Chabaux et al., 2003). Since the interaction Zhang and Schaap, 2019). An extreme example is represented by karstic between fresh and saline groundwater in the subterranean estuary is also or volcanic coastal aquifers, where localized hydraulic conductivity can often accompanied by a redox gradient (Charette and Sholkovitz, 2006; reach values of ~107 cm⋅d− 1 (Li et al., 2020). On the other hand, McAllister et al., 2015), the dissolution of hydrous oxides under physical mechanisms driving groundwater flow also control the reducing conditions may increase the activity of Ra isotopes in groundwater velocity (as well as the spatial scale of the process), and groundwater (Beneš et al., 1984; Gonneea et al., 2008). Conversely, therefore the groundwater transit times. In coastal aquifers, these mineral precipitation processes can remove Ra from groundwater. Due driving forces include the terrestrial hydraulic gradient and their sea­ to the very low molar activities of Ra in groundwaters, the removal of Ra sonal oscillations, shoreface circulation and tidal pumping, wave setup from groundwater usually occurs by co-precipitation with other phases. and wave pumping, bioirrigation, flow- and topography-induced Ra2+ may co-precipitate with Mn and Fe hydrous oxides under oxidizing advection, among others (Santos et al., 2012). These forces can be and high pH (>7) conditions (Gonneea et al., 2008; Porcelli et al., 2014), linked to the five SGD pathways outlined in Section 3. Obviously, as well as with sulfates [e.g., (Ba, Ra, Sr)SO4] or carbonate [(Ca, Ra) groundwater transit times are not only determined by the driving CO3] (Kiro et al., 2013; Kiro et al., 2012; Porcelli et al., 2014). However, mechanism itself, but also by the intensity and frequency of the physical due to the relatively long time scales of weathering and precipitation forces (e.g., wave and tidal frequency and amplitude, magnitude and processes, these vectors are often considered to exert negligible controls seasonality of the hydraulic gradient, recurrence and intensity of strong on the activities of 224Ra, 223Ra and 228Ra in coastal groundwaters, episodic wave events, etc.) (Rodellas et al., 2020; Sawyer et al., 2013). although they might be relevant for 226Ra (Porcelli and Swarzenski, The common way to evaluate the effect of transit time on the degree 2003). of enrichment of Ra isotopes in groundwater is usually performed using 1-D advective-dispersive solute transport models. These models 4.4. Groundwater transit time commonly assume: i) steady state aquifer conditions (i.e., activities do not vary with time; dA/dt = 0); ii) that the effects of dispersive and The last fundamental factor that regulates the activities of Ra iso­ diffusive transport in relation to advective transport are negligible, and topes in aquifers is the time of groundwater to travel along a certain flow iii) that the exchange between the solid surfaces and the solution prin­ path, commonly known as groundwater transit time, which also repre­ cipally occurs via ion exchange, neglecting the processes of weathering sents the time that the groundwater is in contact with the solids in the or precipitation (Kiro et al., 2012; Michael et al., 2011; Tamborski et al., aquifer (Diego-Feliu et al., 2021; Rajaomahefasoa et al., 2019; Vengosh 2017a; Diego-Feliu et al., 2021). The underlying assumptions of these simple models may not be valid for all groundwater systems, although they allow understanding of Ra distribution within aquifers. Based on these models, the activities of Ra isotopes in groundwater increase along a flow path towards an equilibrium with Th activities in the aquifer solids (i.e., Th activities in the alpha recoil zone). This equilibrium, defined as bulk radioactive equilibrium by Diego-Feliu et al. (2021), is reached when the activities of Ra do not vary along the groundwater flowpath (dA/dτ = 0). The characteristic groundwater transit time needed to reach equilibrium between the isotopes of Ra and Th depends on the distribution coefficient of Ra (KD) as well as on the decay constant of each Ra isotope (production and decay) (Beck et al., 2013; Michael et al., 2011). The Ra loss due to advective transport when Ra is pre­ dominantly desorbed (low KD; e.g., ~10− 2 m3⋅kg− 1) is higher than when Ra is mostly adsorbed onto grain surfaces (high KD; e.g., ~102 m3⋅kg− 1). Consequently, the characteristic groundwater time to reach the equi­ librium with Th isotopes decreases as the distribution coefficient of Ra increases (Fig. 5). If Ra was completely desorbed (e.g., KD ~ 0 m3⋅kg− 1), the time required to produce 50% of the equilibrium activity of a given Ra isotope would be equal to its half-life. However, in natural envi­ ronments (KD ~ 10− 3 to 103 m3⋅kg− 1; Beck and Cochran, 2013), this activity would be reached within a significantly shorter time, ranging Fig. 5. Transit time of groundwater with aquifer solids required to produce from minutes to hours for the short-lived Ra isotopes, from hours to days 50%, 90% and 99% of equilibrium activity for each Ra isotope as a function of for 228Ra and from days to years for 226Ra (Fig. 5). Therefore, the the distribution coefficient (KD) and the retardation factor (RRa) of Ra (see enrichment rate of Ra isotopes strongly depends on the characteristics of definition in 8.1). Results are derived from a one-dimensional model formulated in Diego-Feliu et al. (2021) and Michael et al. (2011). Distance is also shown the coastal aquifer (e.g., the higher the salinity, the longer the time to considering a groundwater velocity of 1 cm⋅d− 1. Notice that the scale of the Ra reach equilibrium concentrations). The relative difference between the distribution coefficient and the retardation coefficient are logarithmic. Notice characteristic transit time of each isotope is equivalent to the ratio of that the 50% line of 223Ra overlays the 90% line of 224Ra. their decay constants. For example, the transit time needed for 228Ra to 7 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Fig. 6. The spatio-temporal scales of different SGD pathways considering a groundwater advection rate of 1 cm d− 1. Qualitative boundaries for the different pathways are based on common spatial scales and KD of the different pathways (note that R = 1 + KD ρb/Ф). The transit time required to produce more than 50% of the different Ra isotopes is also indicated (solid lines represent 50% of equilibrium; see Fig. 5). reach equilibrium with 232Th is ~570 times higher (λRa− 228/λRa− 224) porewater exchange and shoreface seawater circulation) is not sufficient than that for 224Ra to reach the equilibrium with 228Th. Characteristic to produce measurable long-lived Ra isotopes activities (King, 2012; transit times can be converted to equilibrium distances by assuming a Michael et al., 2011; Moore, 2010; Rodellas et al., 2017). constant velocity of groundwater through the aquifer. Considering a velocity on the order of 1 cm⋅d− 1, all of the Ra isotopes would reach 5. Sources and sinks of Ra isotopes in the water column equilibrium within a few centimeters along the flow path in fresh groundwaters, while lengths would be on the order of tens of cm for The application of Ra isotopes as tracers of SGD in the marine 223 Ra and 224Ra, ~1 m for 228Ra and ~ 500 m for 226Ra in saline environment (at local, regional and global scales) requires a compre­ groundwater. Notice that these lengths are specific for the assumed hensive understanding of their sources and sinks in the water column. As velocity and the distribution coefficients used. For instance, Tamborski summarized in several studies (e.g., Charette et al., 2008; Moore, 2010, et al. (2019) showed that short-lived Ra isotopes reached secular equi­ Moore, 1999; Rama and Moore, 1996; Swarzenski, 2007), the most librium after a flowpath of several meters in a sandy barrier beach from a common sources of Ra isotopes to the ocean can be classified into six hypersaline coastal lagoon, when groundwater velocities were on the groups: 1) Atmospheric input from wet or dry deposition (Fatm); 2) order 10–100 cm⋅d− 1. Discharge of surface water, such as rivers or streams (Friver); 3) Diffusive Given that physical mechanisms control the groundwater transit fluxes from underlying permeable sediments (Fsed); 4) Input from times in the subterranean estuary, the five pathways of SGD described in offshore waters (Fin-ocean); 5) Production of Ra in the water column from Section 3 are likely to be differently enriched in Ra isotopes, as they have the decay of Th parents (Fprod) and 6) inputs from SGD (FSGD). Major different driving forces and spatio-temporal scales. The enrichment rate sinks of Ra in coastal systems are 1) Internal coastal cycling, which in­ of Ra isotopes in each SGD pathway can be evaluated by comparing the cludes the biological or chemical removal of Ra isotopes through co- ingrowth rates of different Ra isotopes with the most common spatio- precipitation with minerals such as Ba sulfates or Fe (hydr)oxides and temporal scales and groundwater salinities of the different pathways Ra uptake (Fcycling); 2) Radioactive decay of Ra (Fdecay) and 3) Export of (Fig. 6). Assuming a relatively homogeneous geological matrix, Ra offshore (Fout-ocean). A conceptual generalized box model summari­ groundwater salinities are likely to control the KD of the different SGD zing all the pathways of removal or enrichment of Ra isotopes in the pathways. As an orientative threshold, here we consider that a given Ra coastal environment is presented in Fig. 7. isotope becomes significantly enriched along a specific SGD pathway when 50% of its equilibrium activity has been reached (Fig. 6). This 50% 5.1. Radium sources enrichment is an arbitrary (though reasonable) threshold because the ability of tracing Ra inputs from a given pathway depends on several The relative magnitude of the different Ra sources vary according to site-specific factors, such as equilibrium activities, U/Th concentration the characteristics of the study site (e.g., presence of rivers or streams, in the coastal geological matrix, magnitude of SGD flow and in­ characteristics of sediments, water column depth), as well as the half-life terferences from other Ra sources (e.g., sediments, rivers). However, this of the specific Ra isotope used in the mass balance (Fig. 8). The different qualitative comparison allows the identification of those SGD pathways source terms for Ra isotopes into the coastal ocean and their relative that might be significantly enriched in the different Ra isotopes (Fig. 6). importance are described below (excluding the Ra inputs from offshore As illustrated in this assessment, short-lived Ra isotopes may become waters (Fin− ocean), which will be considered in the net Ra exchange with enriched in all of the SGD pathways. On the other hand, the transit time offshore waters (see section 5.2)). within the coastal aquifer of short-scale processes (e.g., cm-scale The input of Ra through atmospheric deposition onto oceans (Fatm) 8 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Fig. 7. Conceptual model illustrating the sources (blue) and sinks (red) of Ra isotopes in the coastal ocean. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) includes dissolved Ra in precipitation and, mainly, desorption from at­ streams following a major rain event (Moore et al., 1998) is an addi­ mospheric dust or ash entering the water column. The highest relative tional input. Anthropogenic activities like channels, oil and gas in­ contribution of Ra from atmospheric sources will be expected to occur in stallations, outfalls and waste water treatment plants may also supply large basins (i.e., the higher the ratio of ocean surface area to coastline Ra to coastal waters, although in most cases they prove to be a minor Ra length, the higher the potential importance of atmospheric inputs is) source (e.g., Beck et al., 2007; Eriksen et al., 2009; Rodellas et al., 2017; and/or in areas affected by large atmospheric inputs (e.g., areas affected Tamborski et al., 2020a, 2020b). by the deposition of large amounts of dust or ash). However, even in Sediments are a ubiquitous source of Ra isotopes (Fsed), but their large areas affected by Saharan dust inputs (e.g., the Atlantic Ocean or relative importance as a Ra source is highly dependent on the charac­ the Mediterranean Sea), the percentage of the atmospheric contribution teristics of the sediment (e.g., sediment grain size, sediment mineral is less than 1% of the total Ra inputs (Moore et al., 2008; Rodellas et al., composition, porosity), the specific Ra isotope considered, the mecha­ 2015a) (Fig. 8). Atmospheric input is expected to be smaller in areas nism that releases Ra from the sediments and the spatial scale of the such as bays or coves, and therefore it is commonly neglected in Ra mass study area (Fig. 8). Mechanisms that may release Ra from sediments, balances (Charette et al., 2008). excluding groundwater flow, are molecular diffusion, erosion, bio­ Surface water inputs (Friver) includes both the fraction of Ra dis­ turbation, or sediment resuspension. High Ra in the porewater results in solved in surface waters and the desorption of Ra from river-borne a molecular diffusion of Ra from the sediments to the water column (e. particles as these particles encounter salty water. Surface water input g., Beck et al., 2008; Garcia-Orellana et al., 2014; Garcia-Solsona et al., includes both rivers and streams discharge and inputs from freshwater 2008). Other processes such as sediment resuspension, diagenesis and and salt marshes, stormwater runoff and anthropogenic sources, such as bioturbation may also enhance the exchange of Ra between sediment discharge from wastewater treatment facilities. There are several studies and the water column (e.g., Burt et al., 2014; Garcia-Orellana et al., that have quantified the contribution of Ra from rivers and streams to 2014; Moore, 2007; Rodellas et al., 2015b; Tamborski et al., 2017c). For coastal and open ocean areas (Krest et al., 1999; Moore, 1997; Moore instance, bioturbation has been suggested to increase the 228Ra flux by a and Shaw, 2008; Ollivier et al., 2008; Rapaglia et al., 2010). Most of the factor of two over the flux due to molecular diffusion only (Hancock studies conclude that the fraction of dissolved Ra in river waters is a et al., 2000). In sediments of the Yangtze estuary, bioirrigation was minor contribution compared with the Ra desorbed from surface water- found to be more important for the 224Ra flux from the sediments than borne particles (Krest et al., 1999; Moore et al., 1995; Moore and Shaw, molecular diffusion and sediment bioturbation (Cai et al., 2014). The 2008). The importance of surface water as a Ra source obviously de­ significance of Ra input from sediments largely depends on the pro­ pends on the presence (and significance) of rivers and streams in the duction time of each Ra isotope relative to the Ra-releasing mechanism. investigated area. The relative contribution of surface water might be Inputs from sediments can typically be ignored for long-lived Ra iso­ significant in estuaries or other areas influenced by river discharge (e.g., topes in small-scale studies due to their long production times (e.g., Beck et al., 2008; Key et al., 1985; Luek and Beck, 2014; Moore et al., Alorda-Kleinglass et al., 2019; Beck et al., 2008; Beck et al., 2007; 1995; Moore and Shaw, 2008) (Fig. 8). Marshes are also commonly Garcia-Solsona et al., 2008). However, in basin-wide or global-scale considered as a major source of Ra due to both erosion of the marsh and studies, the large seafloor area results in long-lived Ra sediment input subsequent desorption of Ra from particles and porewater exchange that might be comparable to SGD input (Moore et al., 2008; Rodellas (Beck et al., 2008; Bollinger and Moore, 1993; Bollinger and Moore, et al., 2015a) (Fig. 8). The fast production of short-lived Ra isotopes in 1984; Charette, 2007; Tamborski et al., 2017c). Runoff or ephemeral sediments, which is set by their decay constants, results in a near- 9 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Fig. 8. Relative contribution of different sources (river, atmosphere, sediments and SGD) to the total 224Ra (upper row) and long-lived Ra isotopes (lower row) for six study sites with distinct characteristics: Huntington Beach, USA (sandy intertidal beach face; Boehm et al., 2004 and Boehm et al., 2006); Bangdu Bay, Korea (semi- enclosed bay on a volcanic island; Hwang et al., 2005); Port of Maó, Spain (semi-enclosed harbor with large resuspension of sediments; Rodellas et al., 2015a, 2015b); La Palme Lagoon, France (micro-tidal coastal lagoon with karst springs; Tamborski et al., 2018); York River estuary, USA (micro-tidal estuary; Luek and Beck, 2014); Atlantic Ocean (upper 1000 m of water column; Moore et al., 2008). The sites are organized according to increasing water residence times of the study area (left to right). The Ra isotope used is indicated in the middle of the pie chart. continuous availability of 223Ra and 224Ra in sediments, which can result 5.2. Radium sinks in comparatively larger fluxes of short-lived Ra isotopes to the water column relative to the long-lived ones. Particularly in areas with coarse- The two main Ra sinks for most of the coastal systems are the decay grained sediments with low Ra availability for sediment-water ex­ of Ra isotopes in the water column (Fdecay) and net Ra exchanged with change, fluxes of longer-lived Ra isotopes from seafloor sediments usu­ offshore waters (Fout− ocean- Fin− ocean). As with the Ra sources, the rele­ ally account for a minor fraction of the total Ra inputs (Beck et al., 2008, vance of the Ra sink terms is largely dependent on the characteristics of Beck et al., 2007; Garcia-Solsona et al., 2008). However, sediments can the water body (mainly water residence time) and on the Ra isotope used turn into a major source of Ra isotopes to the water column in shallow (Fig. 9). Other sinks are grouped together as internal cycling (Fcycling). water bodies, in systems covered by fine-grained sediments (substrate These processes include Ra co-precipitation with salts (e.g. barium sul­ with a high specific surface area), sediments with a high content of U phate) or Fe-Mn (hydr)oxides that occur in estuaries, coastal lagoons or and Th-series radionuclides, and/or in areas affected by processes that polluted areas (e.g., Alorda-Kleinglass et al., 2019; Kronfeld et al., 1991; favor the Ra exchange between sediments and overlying waters, such as Neff and Sauer, 1995; Snavely, 1989), the scavenging of Ra with sinking bioturbation (Cai et al., 2014), sediment resuspension (Burt et al., 2014; particles (Moore et al., 2008; Moore and Dymond, 1991; van Beek et al., Rodellas et al., 2015b) or seasonal hypoxia (Garcia-Orellana et al., 2007), uptake by biota, including the incorporation into calcium car­ 2014). bonate, barium sulphate or calcium phosphate lattice of shells and fish The production of Ra isotopes from their dissolved Th parents bones (Iyengar and Rao, 1990; Szabo, 1967), and the adsorption of Ra to (Fprod) is commonly avoided in SGD studies by reporting Ra activities as the outer surface of algae’s or to their internal non-living tissue com­ “excess” activities in relation to their respective progenitors. This is ponents (Neff, 2002). These various internal cycling processes are usually a minor Ra source because Th is a particle reactive element and generally a negligible Ra sink compared to radioactive decay and ex­ is rapidly scavenged by particles sinking through the water column. The change with offshore waters. However, when using long-lived Ra iso­ amount of 232Th, which is introduced into the water column by disso­ topes and conducting basin-scale and global ocean budgets (i.e., low lution from particles supplied by rivers, runoff, atmospheric deposition decay and low exchange), Fcycling should be taken into account (Moore or resuspension, is very low and therefore water column dissolved 232Th et al., 2008; Rodellas et al., 2015a) (Fig. 9). activities are usually orders of magnitude lower than its daughter 228Ra. The decay of Ra isotopes (Fdecay) is characteristic output term of Dissolved activities of 227Th (and 227Ac), 228Th and 230Th in the water mass balances using radioactive isotopes as tracers. The loss of Ra due to column are also low and thus the production of 223Ra, 224Ra and 226Ra decay is usually a term that is relatively easy to constrain because it only from their decay generally represents a minor source term. depends on the Ra inventory of the study site and the decay constant of Submarine Groundwater Discharge (FSGD) is often a primary source the isotope used. Thus, when decay is a primary sink of Ra in the system of Ra isotopes to the ocean (Fig. 8) and it is the target flux in SGD in­ studied, Ra removal can be accurately contrained provided that the Ra vestigations. This term includes Ra inputs from any water flow across water-column inventory has been adequately determined (Rapaglia the sediment-water interface, which can be supplied through the five et al., 2012). The importance of the decay term will depend on the half- different SGD pathways outlined in Section 3 (Fig. 3). Different SGD life of the isotope used and the water residence time of the system pathways occur at different locations (e.g., nearshore, beachface, studied, as illustrated in Figs. 9 and 10. The loss of Ra due to radioactive offshore) and have different groundwater compositions (e.g., fresh, decay is negligible for long-lived Ra isotopes in coastal areas with brackish or saline groundwater) and characteristic groundwater transit relatively short residence times (< 100 days) (Fig. 9). In regional and times within the subterranean estuary (e.g., hours-days for cm-scale global-scale studies, the decay of 228Ra needs to be considered and porewater exchange and months-decades for terrestrial groundwater usually represents the main sink of this isotope (e.g., Charette et al., discharge). All of them are, however, potential sources of Ra isotopes to 2015; Kwon et al., 2014; Liu et al., 2018; Moore et al., 2008; Rodellas the ocean and thus need to be taken into account when Ra isotopes are et al., 2015a) (Fig. 9). On the contrary, 226Ra can be considered as a used as SGD tracer. stable element in SGD studies because decay is negligible on the time- scale of processes occurring in coastal or regional systems. For the short-lived Ra isotopes, the removal of Ra due to decay must be 10 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Fig. 9. Relative contribution of different sinks (decay, exchange and scavenging) to the total 224Ra (upper row) and long-lived Ra isotopes (lower row) outputs for six study sites with distinct characteristics (see Fig. 8 for site references and details). accounted for in mass balances for most coastal regions (Fig. 9). How­ considered, and subsequently convert it into a volumetric water flow ever, in rapidly flushed systems it may be almost negligible for 223Ra and (SGD; m3 d− 1) by characterizing the Ra activity in the discharging to a lesser extent for 224Ra (i.e., residence times on the order of few groundwater (i.e., the SGD endmember, CRa-SGD; Bq m− 3) (Eq. 1). hours) (e.g., Alorda-Kleinglass et al., 2019; Boehm et al., 2004; Trezzi FSGD et al., 2016) (Figs. 9 and 10). SGD = (1) CRa− SGD The relative contribution of the exchange of Ra isotopes due to the mixing between coastal and open waters (offshore exchange - the There are three basic strategies to quantify total SGD fluxes (FSGD) difference between Fout-ocean and Fin-ocean) depends on the study site. using Ra isotopes: i) mass balances; ii) endmember mixing models and Nearshore waters usually have higher Ra concentrations than offshore iii) offshore flux determination from horizontal eddy diffusive mixing. waters, and thus there is usually a net export of Ra offshore. The loss of The most comprehensive and widely applied approach is the Ra mass Ra due to the mixing between coastal and open waters is directly linked balance, where the flux of Ra supplied by SGD is usually quantified by a to the flushing time of Ra in the coastal system. In open coastal systems “flux by difference approach”, which considers all the potential Ra or sites with relatively short flushing times (e.g., systems with short sources and sinks identified in Fig. 7: water residence times and/or high dispersive mixing with offshore wa­ ∂AV ( ) ters), the export of Ra isotopes offshore is frequently the primary = Fatm + Friver + Fsed + Fprod + Fin− + FRa− (2) ocean SGD ∂t removal term and thus, it is commonly one of the most critical param­ ( ) − Fcycling + Fdecay + Fout− eters to be determined in Ra mass balances, particularly for the long- ocean lived Ra isotopes (Fig. 10) (Tamborski et al., 2020a, 2020b). In semi- where A is the average Ra activity in the study area, V is the volume enclosed or enclosed coastal environments (e.g., bays, coastal lagoons, affected by SGD and t is time (i.e., ∂AV ∂t is the change of Ra activity in the coves), flushing times are usually long, and Ra losses due to mixing with study area over time). This approach is often simplified in coastal areas offshore waters is less important for the short-lived Ra isotopes. How­ by neglecting the commonly minor Ra sources and sinks (atmospheric ever, for the long-lived Ra isotopes this process is the main sink even in inputs, production and internal cycling) and assuming that the system is these semi-enclosed water bodies (Fig. 9), requiring an appropriate in steady state (i.e., ∂AV ∂t = 0), and thus all quantifiable Ra input fluxes are characterization of the mixing term to obtain an accurate quantification of SGD. Different Ra isotopes are commonly combined to constrain subtracted from the total output with the residual being attributed to offshore exchange (e.g., Moore, 2000a, 2000b; Moore et al., 2006) (see SGD: section 8.3), although this output term can also be estimated using other (A − Aocn )V approaches such as hydrodynamic and numerical models (e.g., Chen FRa− SGD = + AVλ − Friver − Fsed (3) tf et al., 2003; Lin and Liu, 2019; Warner et al., 2010), tidal prism (e.g., Dyer, 1973; Petermann et al., 2018; Sheldon and Alber, 2008) or direct where Aocn is the Ra activity of the open ocean water that exchanges current/flow measurements (e.g., Rodellas et al., 2012; Shellenbarger with the study area, tf is the flushing time of Ra in the system due to et al., 2006). mixing and λ is the radioactive decay constant of the specific Ra isotope used. Notice that the first and second terms on the right side of the 6. Quantification of SGD using Ra isotopes equation describe offshore exchange (i.e., Fout-ocean - Fin-ocean) and radioactive decay, respectively. Flushing time is an integrative time 6.1. Ra-based approaches to quantify SGD parameter used to describe the solute (Ra isotopes) transport processes in a surface water body (due to both advection and dispersion) and it Ra isotopes are suitable SGD tracers mainly because i) activities in relates the mass of a tracer and its renewal rate due to mixing (Monsen groundwater are typically 1–2 orders of magnitude higher than in et al., 2002). Notice also that we refer to flushing time of Ra due to coastal seawater; ii) Ra isotopes in coastal areas are usually primarily mixing only, as if it were a conservative and stable solute. Some authors sourced from SGD; iii) they behave conservatively in seawater and iv) use the concept of water residence time (i.e., the time a water parcel they have different half-lives (ranging from 3.7 days to 1600 years), remains in a waterbody before exiting through one of the boundaries; tr) therefore allowing the tracing of coastal processes on a variety of time- to refer to tf, but this term is not appropriate for systems influenced by scales. The common approach to quantify SGD is to quantify first the Ra dispersive mixing (such as many of coastal sites) or evaporation where flux supplied by SGD (FSGD; Bq d− 1), regardless of the SGD pathway tracers and water are transported due to different mechanisms. This 11 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 mixing (focn), the SGD source 1 (fSGD1) and the SGD source 2 (fSGD2), and the Ra isotope endmembers are identified by the same series of sub­ scripts. With this approach, the fSGD can be determined for each location where there is a Ra isotope measurement (22XRa). Short-lived Ra iso­ topes can also be used in the equations by incorporating their decay term. This approach is particularly useful to resolve changes in mixing between different Ra sources over tidal time-scales, e.g., within marsh creeks (Charette, 2007). The fractional SGD contributions derived from this mixing model can be converted into volumetric fluxes by consid­ ering the time scales of water mass transport (tr) in the system under study, as follows: V⋅fSGD SGD = (7) tr If the water outflow of the system can be directly measured (e.g., using an Acoustic Doppler Current Profiler – ADCP), then the end­ member fraction can be directly multiplied by the measured flow (e.g., Rodellas et al., 2012; Tamborski et al., 2021). The third approach is based on using offshore transects of short-lived Ra isotopes to estimate coastal mixing rates via the offshore coefficient Fig. 10. Relative contribution of the decay term in Ra mass balances as a of solute dispersivity, Kh [m2⋅s− 1], which can be used in conjunction function of flushing time (d) in a study area. Notice that the scale of the flushing with the offshore gradient of long-lived Ra isotopes to estimate the time is logarithmic. offshore export of 228Ra or 226Ra (Moore, 2015; Moore, 2000b). Several assumptions are required to apply this model: i) there is no additional approach is commonly applied in environments where the distribution input of Ra beyond the nearshore source, ii) the system is in steady state of Ra isotopes in the study site is influenced by multiple Ra sources and on the timescale of the isotope used (see section 7.3), iii) advection is sinks (Fig. 7). The accuracy of the estimated Ra flux supplied by SGD will negligible in any direction; iv) the open ocean Ra isotope activities are thus largely depend on the accuracy with which the most relevant negligible and v) a flat seabed or the presence of a stratified water col­ sources and sinks of Ra are determined (Rodellas et al., 2021). Indeed, umn of a constant thickness (Knee et al., 2011; Moore, 2015; Moore, given that SGD is quantified by a difference of fluxes, large uncertainties 2000b). Considering these assumptions, the measured log-linear are often associated with the SGD estimates in systems where SGD fluxes decrease in the activity of 223Ra or 224Ra offshore can be used to only represent a minor fraction of total inputs. Most studies quantify determine Kh from the following simplified advection-diffusion equation total SGD with a mass balance by using a single Ra isotope and use other (Eq. 8). short-lived isotopes to estimate the flushing time (tf) (see section 8.3) √̅̅̅̅̅̅̅ that is required in SGD estimations (Eq. 3) (e.g., Gu et al., 2012; Kim Ax = A0 e− x λ/Kh (8) et al., 2008; Krall et al., 2017). Ra isotopes can also be combined through concurrent mass balances to simultaneously quantify SGD derived from where A0 and Ax are the Ra activities at the coast and at a distance x from different aquifers or pathways (Rodellas et al., 2017). the coast, respectively. A detailed discussion on the model and its as­ The second approach estimates SGD fluxes using mixing models sumptions is included in Moore (2015, 2000b). The mixing coefficient between different endmembers (e.g., Charette and Buesseler, 2004; Kh derived from the logarithmic-linear scale plot of the short-lived Ra Charette, 2007; Moore, 2003; Young et al., 2008). This approach is isotopes activities is then combined with the offshore gradient of long- essentially a simplification of the mass balance approach where all Ra lived Ra isotopes (linear-linear decrease) to estimate the export inputs are attributed to water flows (e.g., groundwater, rivers) and offshore of 228Ra or 226Ra. This long-lived Ra flux offshore needs to be where the mixing offshore is assumed to represent the main Ra sink balanced by Ra inputs from all the potential sources, and thus estimating (Moore, 2003). This approach is thus recommended for environments SGD from this approach requires subtracting the contribution of all the where mixing between SGD and the ocean controls the distribution of Ra sources aside SGD from the estimated flux offshore (similar to the mass in the study site and where other Ra sources (e.g., sediment inputs) and balance approach) (Moore, 2015, Moore, 2000b). This approach is sinks (e.g., decay) can be neglected. If various Ra isotopes are used, this recommended for open coastal systems where the export of long-lived method can also be used to distinguish the relative contribution of SGD Ra isotopes offshore due to diffusive mixing is frequently the primary and other water sources (e.g., rivers; Dulaiova et al., 2006a, 2006b) or to removal term for 226Ra and 228Ra (e.g., Bejannin et al., 2020; Boehm separate different SGD components (e.g., SGD fluxes from confined and et al., 2006; Dulaiova et al., 2006a, 2006b; Rodellas et al., 2014). One of unconfined aquifers (Charette and Buesseler, 2004; Moore, 2003), the advantages of using this approach is that it allows combining short- terrestrial SGD and marsh porewater (Charette, 2007). For example, one lived Ra estimates of Kh with solute gradients offshore to obtain the rate can consider a system with two Ra sources (e.g. SGD1 and SGD2), aside of solute transport offshore (e.g., Charette et al., 2007). For these rea­ from offshore exchange. SGD inputs can be estimated by defining the sons, this approach has been widely applied in oceanographic studies major contributors to the Ra isotope budgets of two isotopes (for focused on the transport of solutes supplied by a combination of example, 226Ra and 228Ra) and solve a series of simultaneous equations boundary exchange processes and is generally not applied to discrimi­ (Moore, 2003). nate different Ra sources (e.g., SGD, rivers, sediments) (e.g., Charette et al., 2016; Jeandel, 2016; Vieira et al., 2019, 2020). focn + fSGD1 + fSGD2 = 1 (4) 228 Raocn ⋅focn + 228 RaSGD1 ⋅fSGD1 + 228 RaSGD2 ⋅fSGD2 = 228 Ra (5) 6.2. Determination of the Ra concentration in the SGD endmember 226 Raocn ⋅focn + 226 RaSGD1 ⋅fSGD1 + 226 RaSGD2 ⋅fSGD2 = 226 Ra (6) The above-mentioned approaches of SGD quantification rely on an appropriate characterization of the Ra activity in the discharging where 228Ra and 226Ra represent the measured Ra activity for a given groundwater (i.e., the SGD endmember) (Eq. 1). This characterization time period, f is the water mass fraction contributed by coastal ocean and selection of the Ra activity in the discharging groundwater often 12 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 represents the main source of uncertainty for the final SGD estimates, (Gonneea et al., 2008; Moore, 1999; Rocha et al., 2021), that can since Ra activities in the coastal aquifer can vary significantly (1–3 or­ further complicate differentiation between the different groundwater ders of magnitude) over space and time depending on the geochemical discharge processes. To simplify this approach, previous studies have and hydrogeological characteristics at each study site (e.g., Cerdà- argued that some of these pathways are negligible either because the Domènech et al., 2017; Cho and Kim, 2016; Cook et al., 2018; Duque driving forces are not significant (e.g., no tidal-driven recirculation in et al., 2019; Gonneea et al., 2008, 2013; Michael et al., 2011). Several microtidal environments; Alorda-Kleinglass et al., 2019; Krall et al., studies highlighted the inherent difficulties and uncertainties related to 2017; Rodellas et al., 2017; Trezzi et al., 2016) or because low per­ the selection of representative SGD endmembers (Cerdà-Domènech meabilities of ocean sediments restrict inputs from some SGD path­ et al., 2017; Cho and Kim, 2016; Cook et al., 2018; Gonneea et al., 2013; ways (e.g., Beck et al., 2007; Michael et al., 2011; Michael et al., 2005). Michael et al., 2011). There is no general framework to characterize SGD Some authors also identified groundwater pathways, which were endmembers valid for all study sites. The SGD endmember is commonly weakly enriched in the specific Ra isotope used, and then argued that determined from near-shore piezometers or inland wells (e.g., Charette the system was not significantly influenced by Ra inputs from this et al., 2013; Garcia-Solsona et al., 2010a; Tovar-Sánchez et al., 2014), pathway (e.g., Beck et al., 2007; Garcia-Solsona et al., 2008; Moore porewaters collected at the seafloor, usually within the first meter below et al., 2019). It should be additionally noted that different SGD path­ the sediment-water interface (Rodellas et al., 2017; Tamborski et al., ways often mix before discharge and do not occur as isolated processes. 2018), direct measurements of SGD actually discharging to the system When two pathways mix before discharge, this mixed SGD can be through seepage meters or springs (Garcia-Orellana et al., 2010; Montiel treated as a single pathway, provided that the Ra concentration in this et al., 2018; Weinstein et al., 2007) or laboratory experiments to esti­ mixed flow is appropriately characterized (Cook et al., 2018). mate the Ra activity in equilibrium with sediments (Beck et al., 2008; Garcia-Solsona et al., 2008). Once groundwater samples are collected at 7. Improving the current use of Ra isotopes as SGD tracers different locations and/or times, in most published studies Ra values are averaged to obtain a ‘representative average’ SGD endmember (e.g., Ra isotopes have been widely applied to quantify SGD in all kind of Beck et al., 2007; Kwon et al., 2014; Lee et al., 2012; Rapaglia et al., locations worldwide and in a large range of physical environments (e.g., 2010). lagoons, wetlands, estuaries, bays or ocean basins). The estimates of An important consideration when determining the SGD endmember SGD derived from Ra isotopes are based on conceptual assumptions that is that the selected Ra concentration needs to be representative of the are often not properly justified or validated, and may involve large discharging groundwater. In coastal systems with a dominant pathway, uncertainties that are usually not quantificable and could lead to inac­ the Ra endmember concentration should thus be representative of this curacies or unrealistic of the SGD estimates. In this section, we discuss pathway. Whilst this recommendation is self-evident, many studies have some key considerations with the aim of improving future applications overlooked or they assumed that the sampled endmember is represen­ of Ra isotopes as SGD tracers. tative for the SGD pathways. Given the general and inherent difficulties of sampling SGD, a common strategy to obtain SGD endmembers is 7.1. Validation of model assumptions and consideration of conceptual measuring Ra concentrations in existing wells or piezometers, which are uncertainties generally installed in the freshwater (or brackish) zone of the coastal aquifer. However, these wells may not span the range of Ra activities The magnitude of SGD can be quantified using a variety approaches, along the salinity gradients in a subterranean estuary and thus they are including field methods (e.g., seepage meters, natural tracers and not necessarily representative of the SGD endmember in sites dominated piezometric level measurements) or computational approaches (e.g., by saline SGD (Pathways 2 to 5) (Michael et al., 2011). Cho and Kim water budget models and numerical solutions of the groundwater flow (2016) highlighted and discussed this issue at a global scale and showed equation) (Taniguchi et al., 2019). Each of these approaches is prone to that the endmembers frequently used to estimate total SGD are often low certain biases and is limited in both the spatial coverage and the SGD salinity groundwater samples (i.e., with relatively low Ra concentra­ pathways it is capturing (Burnett et al., 2006; Zhou et al., 2018). Using tions), even if SGD fluxes are composed mainly of circulatated seawater different approaches can yield order-of-magnitude differences in SGD at (Pathways 2 to 5). These authors re-evaluated previous SGD estimates individual locations and, thus, the selection of the method itself con­ for the Atlantic Ocean and for the global ocean by using only end­ stitutes the first source of uncertainty in any SGD study (Rodellas et al., members with salinities ~ > 10 and showed these fluxes were likely 2021; Zhou et al., 2018). overestimated two- to three-fold only because non-representative low- Tracer approaches are based on conceptual models that represent an salinity groundwaters were included in the determination of the SGD abstraction of a complex and dynamic natural system. Thus, a second endmember (Cho and Kim, 2016). On the other hand, if SGD mainly additional source of uncertainty is linked to the conceptualization of the consists of seawater circulation with short transit times within the natural system, the so-called conceptual or structural uncertainty subterranean estuary (e.g., shoreface seawater circulation or cm-scale (Regan et al., 2002). In the case of Ra isotopes, there are three models (or porewater exchange), the long-lived isotope activities (e.g., 228Ra) of concepts) commonly applied to quantify SGD (see section 6.1): “Mass circulated seawater (Pathways 2 to 5) may be much lower than terres­ balance model” and “mixing model”, which are both based on Ra bud­ trial water activities (even in fresh water of Pathway 1). This bias to­ gets using lumped parameter models, and a “offshore export model”, wards on-shore borehole/well data (long transit times) may actually which is based on a simplified analytical solution to describe Ra trans­ result in an underestimation of SGD, in particular in basin-scale studies. port. Each one of these models is extremely sensitive to its specific In more complex settings where several SGD pathways coexist, the simplifications, assumptions and boundary conditions used to approxi­ SGD endmember should account for the different Ra enrichments in mate the real system. For instance, Ra “mass balances” are strongly the SGD pathways that occur in the study area and should reflect the influenced by the sources and sinks of Ra included in the model (e.g., relative contributions of these different pathways. Obtaining the some studies assume that all Ra inputs can be directly attributed to SGD) representative Ra endmember concentration thus requires, in addition and, mainly, by the parameterization of the different input and output to determining the Ra activity in each pathway, constraining the terms and their intrinsic assumptions. Indeed, the use of different relative contribution of each SGD pathway. This is difficult and re­ methods or parameterizations to quantify the same term in the mass quires previous knowledge of the fluxes from different pathways, balance (e.g., different approaches to estimate offshore exchange rates, which is usually not available (Cook et al., 2018; Michael et al., 2011). sediment diffusion or to produce an averaged Ra concentration in the In addition, when seawater and groundwater mix, there are salinity study area) may lead to order-of-magnitude discrepancies in Ra-derived gradients and redox interfaces, which promote geochemical reactions SGD estimates (Rodellas et al., 2021). The mass balance approach also 13 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 assumes that the integrative parameters can appropriately represent the system (e.g., the model assumes that offshore exchange occurs with an average Ra concentration). Similarly, the “offshore export” model is essentially a simplification of the 1-D advection-dispersion equation, which is very sensitive to the boundary conditions and basic assump­ tions. These assumptions may not be correct in environments with along-shore and across-shore currents (advection cannot be neglected), with Ra inputs occurring beyond the nearshore source, with not uniform vertical profiles of Ra isotopes or non-negligible Ra concentrations offshore, or in systems with a sloped seabed (e.g., Hancock et al., 2006; Lamontagne and Webster, 2019a, 2019b). Additionally, all the Ra models are often based on the assumption that mixing processes affect all the isotopes alike, either through the estimation of flushing times or water ages (tf or tw) using pairs of isotopes (see section 8.3) or by using an offshore diffusivity coefficient (Kh) derived from one isotope to evaluate the transport of another one (Lamontagne and Webster, 2019b). Since Ra transport offshore is often controlled by dispersion, which integrates processes operating at different temporal and spatial scales, mixing parameters (e.g., tf or Kh) are scale dependent and should be independently evaluated for each tracer (Lamontagne and Webster, 2019b; Moore, 2015; Okubo, 1976). Finally, and most importantly, Fig. 11. Average residence times for different Ra isotopes as function of flushing time due to mixing (based on Rodellas et al., 2021). almost all studies assume that the system is in steady state and that all model parameters are constant on the timescale the tracer resides in the system, assumptions that are very often not met – in particular for long- timescales of transport processes exceeds that of radioactive decay (1/tf lived isotopes – and that should be carefully evaluated (see section 7.2). >> λ), the Ra residence time in the system is comparable to the temporal The reliability of Ra approaches to quantify SGD may thus be limited scale of coastal mixing. Thus, the assumptions of constant fluxes and and scientists should be aware of these limitations and try to validate steady state need to be validated on timescales of transport processes their assumptions and constrain the uncertainties of the estimated flows. (Moore, 2015; Rodellas et al., 2021). For systems with negligible mixing Overcoming these limitations is not straightforward and there is not a processes relative to the isotope half-life (1/tf << λ), the Ra temporal general framework for accounting for these conceptual uncertainties or scale is controlled by the radioactive decay of each isotope. In this latter validating assumptions. Some authors have suggested evaluating the case, the assumption of steady state and constant fluxes need to be most sensitive parameters using multiple approaches and obtaining final validated on a timescale comparable to the mean life of the specific SGD fluxes as an ensemble of results (e.g., Rodellas et al., 2021), isotope used. Validating this assumption might thus be particularly combining multiple independent methods to constrain the magnitude of challenging for long-lived Ra isotopes in systems with long flushing SGD (e.g., Zhou et al., 2018) or developing more complex transport times. In those dynamic systems that present a relevant temporal vari­ models (e.g., Lamontagne and Webster, 2019b). For example, ability on the timescale of the tracer used (e.g., systems with variable combining ‘terrestrial’ hydrogeological, oceanographic or geophysical surface water inputs, temporal changes of SGD or variable mixing approaches with tracers tools may help to improve SGD estimates. offshore), the evolution of Ra concentration in sources, sinks and in­ Producing more robust estimates of SGD is clearly a major open research ventories in the water column need to be understood. Neglecting the topic in SGD and Ra investigations (see section 9.2). transience of these parameters could lead to significant errors on the final SGD estimates. In such non-steady state systems, a dynamic modelling approach might provide a better representation of the studied 7.2. The assumption of steady-state needs to be validated system (Gilfedder et al., 2015). A key general assumption in all the Ra-based approaches to quantify 7.3. Ra isotopes cannot be used indistinctly to trace SGD SGD is that all Ra fluxes (e.g., SGD-driven Ra flux, Ra diffusion from sediments, radioactive decay, Ra export offshore) are constant with Ra isotopes have been commonly used to estimate total SGD, often respect to the timescale of the Ra isotope used and that the system is in without consideration as to whether the Ra isotope used is actually steady state (i.e., Ra inputs equal Ra outputs). Therefore, any appro­ capturing the entire discharge pathway(s) underlying these flows. This priate SGD study should validate these assumptions considering the time raises the question of whether or not the four Ra isotopes provide the the tracer resides in the system, i.e. the tracer residence time (Rodellas same SGD estimates and if they therefore can be used indistinctly. et al., 2021). The average residence time of Ra isotopes in the surface In an idealized case where there is a unique SGD pathway or source water system depends on the rate of removal of the isotope and thus on (e.g., a karstic or volcanic area with a main coastal spring that dominates their different sinks. Given that the residence time of Ra isotopes (tRa) in over any diffuse, non-point source discharge), all Ra isotopes should most of the systems is mainly controlled by radioactive decay and export yield the same SGD flux, provided that i) all the Ra sources (aside from offshore (see Fig. 9), it can be estimated following Eq. 9 (Rodellas et al., SGD) and sinks are appropriately constrained, ii) groundwater is suffi­ 2021): ciently enriched with the Ra isotope used as a tracer, and iii) the SGD tRa ≈ 1 (9) endmember is properly characterized. However, most natural systems λ + t1f do not satisfy these idealized conditions because of the ubiquitous presence of multiple driving forces (e.g., hydraulic gradient, wave and where λ is the radioactive decay constant of the specific Ra isotope used tidal pumping, bioirrigation) that result in concurrent inputs of Ra from and tf is the flushing time in the system due to mixing. This flushing time different SGD pathways (Fig. 3). As illustrated in Fig. 6, the different is equivalent to mean water residence time in systems where the influ­ SGD pathways are likely to have different groundwater composition and ence of evaporation and dispersion are negligible. Characteristic Ra characteristic groundwater transit times, resulting in isotope-specific Ra residence time for study sites with different mixing timescales (i.e., enrichments, which may differ by orders of magnitude among different different flushing times) are shown in Fig. 11. In systems where the SGD pathways (Diego-Feliu et al., 2021; Michael et al., 2011). Different 14 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Ra isotopes are thus incorporating different pathways in a characteristic including both water exchange at the boundaries of the system and the proportion, and thus they are likely to produce different SGD estimates. Ra endmember concentration (Tamborski et al., 2020a, 2020b). Given For instance, whilst SGD estimates based on 223Ra and 224Ra are likely to that long-lived Ra isotopes are highly sensitive to mixing in coastal areas incorporate inputs from all the pathways, long-lived Ra isotopes may not (i.e., estuaries, coastal lagoons), they may not be adequate tracers of incorporate SGD induced by short-scale processes (cm-scale circulation SGD in coastal environments when mixing is uncertain (Ku and Luo, or shoreface saline circulation) (King, 2012; Michael et al., 2011; Moore, 2008; Rutgers van der Loeff et al., 2018). Short-lived Ra isotopes are 2010; Rodellas et al., 2017) (Fig. 6). Therefore, when SGD is discharging ideal for coastal areas where timescales are on the order of days; how­ via multiple pathways, all four Ra isotopes cannot be used indistinctly to ever, sediment fluxes must be adequately characterized because they obtain total SGD. often represent a major source of short-lived Ra isotopes (Burt et al., Most Ra-based SGD studies aim at obtaining a single estimate of the 2014; Rodellas et al., 2015b). Water column inventories of short-lived total magnitude of SGD. In these cases, SGD is often treated as a single Ra isotopes integrate over time-scales similar to the Ra isotope half- source, regardless of the difference in Ra activities of the multiple life, and therefore multiple samplings are necessary if seasonal or pathways, and only a single endmember is determined to represent the annual variations in SGD want to be captured. ‘average’ Ra activity in groundwater discharging into the study site (i.e., CRa-SGD in Eq. 1). However, this Ra endmember should not be obtained 7.5. Ra isotopes can contribute identifying the pathways of SGD from simple averages of all the SGD samples collected, because of possible biases towards the most sampled pathway. Michael et al. (2011) A proper evaluation of SGD requires the understanding of the illustrated this issue by comparing SGD estimates obtained by sampling dominant SGD pathways. There are several hydrogeological, geophys­ different pathways and showed that this could produce differences in ical and geochemical characterization techniques (e.g., Folch et al., final SGD estimates by more than an order of magnitude. Therefore, 2020; Gonneea et al., 2008; Swarzenski et al., 2007b; Swarzenski et al., unless there is a comprehensive understanding of the system that allows 2006; Zarroca et al., 2014), as well as numerical modelling approaches constraining the relative contribution of different pathways, SGD esti­ (e.g., Amir et al., 2013; Anwar et al., 2014; Danielescu et al., 2009), that mates derived from a single Ra isotope might not be appropriate at sites provide fundamental information about the origin and spatio-temporal with multiple pathways. scales of the different SGD pathways through the subterranean estuary The combination of multiple Ra isotopes can be instrumental in those (Fig. 6). The characterization of Ra isotopes and especially their activity systems where several pathways are hypothesized to significantly ratios (e.g., 228Ra/226Ra and 224Ra/228Ra) in the subterranean estuary contribute to Ra budgets (see also section 7.5). Both total SGD and the can also provide very useful information to constrain the presence of relative contribution of different SGD pathways can be accurately ob­ different SGD pathways in the study site. For instance, since 228Ra and 226 tained from the concurrent application of multiple Ra isotopes. The Ra belong to different decay chains (U and Th), the activity ratio of 228 number of Ra isotopes to be used in the models will depend on the Ra/226Ra may provide information about the Th/U ratio in the host number of SGD pathways to be determined, as well as the amount of material and thus be used to identify and distinguish groundwater other potential unknowns (e.g., residence time, river contribution). This inflowing from different geological systems (e.g., Charette and Buess­ approach can be applied through concurrent mass balances (see section eler, 2004; Moore, 2006a, 2006b; Swarzenski et al., 2007a, 2007b). 6.1; e.g., Rodellas et al., 2017; Tamborski et al., 2017a) or mixing Additionally, the 224Ra - 228Ra parent-daughter relationship enables the models (e.g., Charette, 2007), provided that the Ra isotopes used have use of the 224Ra/228Ra activity ratio for characterizing different SGD distinct relative enrichments (i.e., distinct Ra isotopic ratios) in the pathways. Whilst short-time-scale SGD pathways are likely to be more different pathways. Ra isotopes can also be combined with other tracers enriched in short-lived Ra isotopes relative to long-lived ones (Fig. 6) (e.g., silica, radon, salinity, stable isotopes, etc.) (e.g., Burnett et al., presenting thus high 224Ra/228Ra activity ratios (> 1), long-scale path­ 2006; Garcia-Orellana et al., 2010; Oehler et al., 2019; Schubert et al., ways are likely to present similar 224Ra and 228Ra activities (i.e., 224 2015) or approaches (e.g., seepage meters, Darcy estimates, heat Ra/228Ra ~ 1) (Diego-Feliu et al., 2021). tracing, tidal prism, hydrogeological models) (e.g., Garcia-Solsona et al., Moreover, the activity ratios of Ra isotopes in coastal seawater can 2010a,b; Povinec et al., 2012; Prieto and Destouni, 2005; Rosenberry also be used to evaluate the predominant discharge pathways of SGD, et al., 2020), which can decisively contribute towards the quantification provided that Ra is mainly supplied to the coastal ocean by SGD (i.e., of the different pathways and the appropriate characterization of SGD. absence of any significant source such as diffusion from sediments, seawater or rivers). In this case and if radioactive decay is accounted for, 7.4. The Ra isotope(s) used need to be carefully selected the activity ratios of coastal seawater must reflect the combined contribution of the different SGD pathways, being biased towards the The selection of the tracer used should ideally be based on the pro­ more relevant ones (Boehm et al., 2006; Moore, 2006b; Rodellas et al., cess to be studied and the characteristics of the study site, rather than on 2014). For instance, at a site dominated by short-scale pathways (e.g., the methods available (e.g., salinity or radon are much better SGD porewater exchange), the 224Ra/228Ra activity ratio of coastal sea water tracers than radium in certain systems). Likewise, the best suited Ra is expected to be comparable to that of porewater (i.e., 224Ra/228Ra > 1). isotope for any specific study also needs to be chosen based on the target SGD pathway of interest, the area studied and according to the sensi­ 7.6. Estimates of SGD-driven solute fluxes need to account for the tivity of the final estimates of Ra inputs and outputs, which is specific for variable composition of different pathways each coastal water system (Tamborski et al., 2020a, 2020b). In principle, the time scale of the half life of the chosen Ra isotope should scale with The most commonly applied method to estimate fluxes of dissolved the size and time escales of the system, such that the short-lived Ra chemicals (e.g., nutrients, metals, contaminants) driven by SGD is by isotopes are most effective at nearshore and embayment scales (e.g., multiplying either the (Ra-derived) SGD water flow by the average Boehm et al., 2004; Knee et al., 2010; Rapaglia et al., 2012), while long- concentration of the chemical solutes in the SGD endmember or the lived Ra isotopes are most effective at regional and global-scales (water SGD-driven Ra flux by the average solute/Ra ratio in the SGD end­ residence times >100 days; e.g., Charette et al., 2015; Kwon et al., 2014; member (Charette et al., 2016; Hwang et al., 2016; Santos et al., 2011; Liu et al., 2018; Moore et al., 2008; Rodellas et al., 2015a). In either Tovar-Sánchez et al., 2014). This straight-forward approach might be scenario, radioactive decay can be easily estimated from the Ra in­ appropriate for systems with one dominant SGD pathway (provided that ventory in the water column and this facilitates characterizing total the endmember concentration is appropriately determined), but it output fluxes in systems where radioactive decay is the primary Ra sink. should not be directly applied in systems with multiple SGD pathways or On the contrary, mixing losses of Ra isotopes can be highly uncertain, sources. This is mainly because the biogeochemical composition of 15 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 discharging fluids from different SGD pathways may be considerably Section 5, include production, decay, desorption, adsorption, weath­ different depending on the origin of solutes, their transformations ering and precipitation, as well as dispersion. The most common analysis within the subterranean estuary and the temporal and spatial scales of of Ra distribution in groundwater often utilizes a one-dimensional the different pathways (e.g., Rodellas et al., 2018; Santos et al., 2012; transport model (Eq. 10), Tamborski et al., 2017a; Weinstein et al., 2011). Therefore, as discussed ∂ARa for Ra isotopes, the solute concentration obtained from averaging = L(ARa ) − λRRa ARa + λP − λВ + λΓ (10) ∂t groundwater samples (and thus the flux estimates) is likely to be biased towards the most sampled pathway, and is not representative of the where ARa [Bq m− 3] is the activity of Ra in groundwater, t [s] is time, λ relative SGD flux from that pathway. For instance, terrestrial ground­ [s− 1] is the decay constant, and RRa [− ] is the linear retardation factor of water discharge (Pathway 1) often contains high concentrations of nu­ Ra. Retardation is defined as the retention of a solute due to interaction trients and other compounds released from anthropogenic activities, with solid phases relative to bulk solution in a dynamic system whereas other SGD pathways may not be influenced by anthropogenic (McKinley and Russell Alexander, 1993). The first term on the right sources. Using only solute concentrations from wells and boreholes hand-side of the eq. (10) is the linear operator for transport (L(ARa) = − located within the low-salinity part of the subterranean estuary may ∇ qARa + ∇ D ∇ ARa, with q [m s− 1] and D [m2 s− 1] as the Darcy flux result in average solute concentrations having a strong anthropogenic vector and the dispersion tensor, respectively) that includes the advec­ signal (e.g., high nutrient concentrations) that might not be represen­ tive and dispersive terms. The second and third terms are the radioactive tative of SGD in sites influenced by pathways other than terrestrial decay of the exchangeable activity of Ra (i.e., adsorbed and dissolved) groundwater discharge (see section 6.2). An appropriate understanding and production from Th in solution, solid surfaces, and in the effective of the magnitude of solute fluxes driven by SGD, requires thus a previous alpha recoil zone (P [Bq m− 3]). The last two terms refer to the rate of co- identification of the dominant SGD pathways, which needs to be based precipitation (B [Bq m− 3]) and weathering (i.e., dissolution) (Γ [Bq on a detailed knowledge of the studied system. Sampling efforts should m− 3]) of Ra, respectively. then be taken to target all the relevant SGD pathways for the particular This equation is often simplified by assuming steady-state conditions system, particularly over the full-salinity gradient of the subterranean (∂ARa/∂t = 0), negligible Ra co-precipitation (В = 0) and dissolution (Γ estuary. By doing so, the conservative and/or non-conservative behavior = 0), negligible dispersion relative to advection (L(ARa) = − ∇ qARa), of the solute of interest can be determined, since some solutes can be and defining a travel time along a streamline (here after transit time, τ chemically modified in the subterranean estuary by varying the ratio of [s]) (Eq. 11; Kiro et al., 2013; Michael et al., 2011; Krest and Harvey, solute to Ra along the SGD flow path. This is an essential step towards a 2003; Schmidt et al., 2011; Tamborski et al., 2017b). correct quantification of SGD-driven solute fluxes in sites with multiple ( ) pathways and thus, towards a comprehensive understanding of the role ARa (τ) = P + ARa,0 − P e− λRRa τ (11) of SGD for coastal biogeochemical cycles. RRa RRa 8. Additional applications of Ra isotopes in groundwater and where ARa, 0 is the activity of groundwater entering a streamline. marine studies This simple solution implicitly assumes that production rates are constant and thus it neglects the effect of spatial and/or temporal vari­ Aside from the application of Ra isotopes to quantify SGD, Ra iso­ ations in the parent isotope activities (e.g., geological heterogeneity). topes can also provide instrumental information that can help to Moreover, the determination of groundwater ages, velocities or transit constrain the different terms needed for SGD evaluations, as well as key times using this equation strongly relies on an accurate determination of information for characterizing hydrological and oceanic systems. In this the retardation factor of Ra (RRa), which may also be spatially and section, we briefly summarize parallel applications of Ra isotopes in temporally variable. The retardation factor of Ra is commonly deter­ groundwater and marine studies, which include: i) assessing transit mined by using the distribution coefficient of Ra (Eq. 12; Michael et al., times in coastal aquifers; ii) estimating solute fluxes across the sed­ 2011). iment–water interface; iii) estimating ages of coastal surface waters; and ρs RRa = 1 + KD (12) iv) quantifying shelf-scale solute fluxes. ϕ 8.1. Assessment of groundwater transit times in coastal aquifers where KD[m3 kg− 1] is the distribution coefficient, ρs [kg m− 3] is the dry bulk mass density, and ϕ is the porosity. Constraining groundwater transit times in a coastal aquifer or in the If the production rate (P), the retardation factor of Ra (RRa) and the subterranean estuary is a key question in characterizing coastal hydro­ inflow boundary activities of Ra (ARa, 0) are known, groundwater geology. Applications include deriving characteristic distances to the transit times can be easily derived from eq. 11 using a single Ra upgradient recharge points, evaluating the connectivity of the coastal isotope. However, determinations of P, RRa and ARa,0 are usually aquifer with the sea, determining ages and velocities of groundwater, difficult and induce large uncertainties that must be propagated in the differentiating variable spatio-temporal scale processes and evaluating calculation of groundwater transit times. In practice, this may be the potential transformation of solutes in the subterranean estuary difficult due to spatial and temporal variability that affects the four Ra (Diego-Feliu et al., 2021; Gonneea and Charette, 2014; Lerner and isotopes differently. Although the processes are integrated along a Harris, 2009; Werner et al., 2013). Since observational techniques are single flowpath, short-lived isotopes have a much shorter ‘memory’ limited in the subsurface, hydrogeological tracers (e.g., H and O stable than long-lived isotopes, so short-lived Ra activities are more sensitive isotopes, 3He, 36Cl, CFCs) are commonly used to obtain key information to production rates and retardation factors more proximal to the on groundwater flow (Leibundgut and Seibert, 2011). In recent years, sampling point. Alternatively, when the model parameters are un­ the activities of Ra isotopes have been used to evaluate groundwater known, groundwater transit times may be determined by using mul­ dynamics in the subterranean estuary (Bokuniewicz et al., 2015; Kiro tiple Ra isotopes or activity ratios (e.g., 224Ra/228Ra) (Diego-Feliu et al., 2013; Tamborski et al., 2019; Tamborski et al., 2017b), which can et al., 2021; Tamborski et al., 2019). When the Ra production rate be a complementary method to other approaches based on atmospheric- changes with depth, a depth-dependent 1D reactive transport model introduced tracers, which mainly decay while traveling through an can be applied (Liu et al., 2019). Non-steady state conditions like aquifer (Cook and Herczeg, 2012). To evaluate the distribution of Ra in variable hydrological forcing, wave set-up as well as tidal fluctuations aquifers, each of the inputs and outputs of Ra to groundwater needs to be may cause non-steady state conditions which may obscure ground­ constrained. These input and output terms, which are summarized in water transit times. Since this model assumes that there is no mixing 16 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 between two different types of groundwater (with different ages and coefficients of dissolved species i and 224Ra, it represents the sum of all Ra content), it can cause problems in the interpretation of the transit processes that affect solute transfer across the sediment-water inter­ times when mixing does occur, largely because mixing is a linear face. The inherent assumption is that molecular diffusion is the rate- process whereas radioactive decay and ingrowth are exponential limiting step for solute transport. This approach is suitable to quan­ (Bethke and Johnson, 2002). Since the determination of groundwater tify fluxes across the sediment-water interface from shallow (< 20 cm) transit times requires some assumptions in solving the transport sediment cores in benthic and coastal (i.e., salt marsh) mud environ­ equation (eq. 11) and involved terms, independent verifications of Ra- ments for oxygen (Cai et al., 2014; Dias et al., 2016), dissolved inor­ derived transit times are strongly recommended. ganic carbon and nutrients (Cai et al., 2015), rare earth and trace elements (Hong et al., 2018; Shi et al., 2018, 2019b). Note that solute 8.2. Solute flux across the sediment–water interface derived from Ra fluxes derived from Eq. 14 integrate over a time-scale of ~10 days and isotopes may greatly exceed fluxes determined from porewater gradients (Cai et al., 2014, 2015) and traditional sediment incubations, which alter The transfer of solutes across the sediment-water interface, through the physical conditions of the sediments, and therefore may not cap­ both diffusion and advection, can have a major effect on the chemical ture small-scale (mm to cm) advective processes (Hong et al., 2018; Shi composition of the overlying waters and shallow sediments. Traditional et al., 2019b). More complicated diagenetic models of 224Ra have been approaches to quantify water exchange and solute fluxes using Ra iso­ developed for systems subject to significant bioturbation and particle topes rely upon one-dimensional reactive-transport models to reproduce reworking, and may thus require parallel sediment measurements of porewater activities as a function of sediment production, bioturbation, excess 234Th to accurately constrain 224Ra fluxes (Cai et al., 2014, molecular diffusion, dispersion and advection (Cochran and Krishnas­ 2015). wami, 1980; Krest and Harvey, 2003; Sun and Torgersen, 2001). Recent Concurrent sediment and water column 224Ra mass balances may be analytical developments have allowed the measurement of surface- used to separate short-scale advective-diffusive processes (i.e., PEX) exchangeable 224Ra and particle-bound 228Th from sediment cores from other SGD pathways (Hong et al., 2017). In deeper sediment sys­ (Cai et al., 2012). Radioactive disequilibrium between the soluble 224Ra tems, where seepage occurs laterally at depth (i.e., as SGD in coarse- daughter and its particle-bound parent 228Th (T1/2 = 1.9 y) may be grained sediments), Eq. 13 can be used for the horizontal (1-D) export produced in sediments and their interstitial pore fluids from the trans­ flux of 224Ra, and thus horizontal water exchange rates (Q; L⋅m− 2⋅d− 1) port of 224Ra as a result of molecular diffusion, bioturbation and irri­ can be estimated (Shi et al., 2019a) using Eq. 16, gation and/or SGD. Assuming that the in-situ decay of 228Th is the sole FRa source of sediment 224Ra, a steady-state mass balance can be written Q= (16) APW − Asea (Cai et al., 2014; Cai et al., 2012) as Eq. 13. ∫∞ ( ) where APW and Asea represents the dissolved 224Ra activity in porewater FRa = λRa ATh,B − ARa,B dz (13) from the seepage layer and seawater, respectively. In turn, solute fluxes may be estimated by considering the concentration difference between 0 where FRa is the flux of 224Ra across the sediment-water interface porewater and seawater of a given solute (Shi et al., 2019a). Similarly, a (Bq⋅m− 2 d− 1), z is the depth in the sediment where disequilibrium occurs two-dimensional advective cycling model can be used to estimate water (m), λRa is the 224Ra decay constant (0.189 d− 1) and ATh,B and ARa,B are exchange and solute fluxes in sandy seabeds subject to wave and tidal respectively the activities of bulk 228Th and 224Ra (Bq⋅cm− 3) of wet pumping (Cai et al., 2020). In conclusion, application of the 224Ra/228Th sediment, which is obtained from a mass unit considering sediment dry disequilibrium method has been significantly expanded upon since its bulk density (ρs = 2.65 g⋅cm− 3) and porosity (ø). A key advantage of the first introduction (Cai et al., 2012), and likely will continue to be 224 Ra/228Th disequilibrium method over traditional reactive-transport expanded upon as more researchers apply this technique to different models (Boudreau, 1997a; Meysman et al., 2005) is that 224Ra produc­ coastal environments. tion is directly measured (via 228Th) for each sediment section, and thus Whereas the approaches outlined above describe fluxes across the this approach accounts for depth-varying production rates. sediment-water interface, diapycnal mixing i.e., the mixing away from Derivation of a benthic 224Ra flux from fine-grained (muddy) sedi­ deep-sea sediments into the interior of the water column is one further ments can in turn be used to quantify the transfer rate (Fi) of a dissolved process which can be quantified using Ra isotopes. As bottom sedi­ species across the sediment-water interface (Cai et al., 2014) using Eq. ments are the source of Ra, the concentration gradient above the 14. bottom sediments can be used to calculate vertical eddy diffusivity (see ⎛ ⎞ Eq. 8). So far very few studies applied this method (Huh and Ku, 1998; ( i ) ∂Ai/ Koch-Larrouy et al., 2015). Nevertheless it may be an important DS ⎜ ∂z ⎟ approach to investigate the fate of benthic sourced solutes in the water Fi = FRa ⎜ ⎝∂ARa/ ⎠ ⎟ (14) DRa S ∂z column. where DiS and DRa S are the in-situ molecular diffusion coefficients for 8.3. Estimation of timescales in coastal surface waters dissolved species i and 224Ra, respectively, corrected for tortuosity and in-situ temperature (Boudreau, 1997b). The concentration gradients of The functioning and vulnerability of a coastal ecosystem (e.g., dissolved species i (∂Ai/∂z) and 224Ra (∂ARa/∂z) are taken as the con­ accumulation of contaminants, risk of eutrophication or harmful algal centration difference between overlying surface waters and shallow blooms) are often closely related to the retention of solutes (e.g., nu­ porewaters, typically measured at 1 cm depth, and are thus net solute trients, contaminants, suspended biomass) and thus the transport fluxes (Cai et al., 2012, 2014). Note that the flux obtained in the Eq. 14 is mechanisms of solutes in the system studied. Therefore, understanding translated from an adjusted Fick’s First Law (Eq. 15). the temporal scales of water and solute fluxes is crucial for coastal ∂c oceanography. Obtaining a flushing time of Ra in a coastal study is also a F = − ξ⋅Φ⋅Ds (15) key parameter needed to estimate SGD through the Ra mass balance or ∂z the mixing model approach (see Section 6.1). Ra isotopes can provide where ξ is an area enhancement factor representing the extended instrumental information in this regard because the different decay subsurface interface and advective influences on the diffusive flux, and constants of each isotope can help infer mixing timescales in coastal Ф is porosity. Although Eq. 14 only includes molecular diffusive systems. There are two basic approaches to estimate water ages using Ra 17 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 isotopes (Moore, 2015): the “mummy” model, appropriate for systems driven by eddy diffusion, whereby the net cross-shelf solute (S) flux can with Ra inputs occurring near the shoreline (Moore, 2000a) and the be linearly scaled with the net cross-shelf 228Ra flux as follows (see also “continuous input” model, recommended for systems where Ra is added 6.1.): over the entire study area, such as a shallow estuary (Moore et al., 2006). ( ) Both approaches are based on using activity ratios of a pair of Ra iso­ 228 Sflux = Raflux × 228 Sshelf − Socean (19) topes to determine water apparent ages, which is the time elapsed since Rashelf − 228 Raocean the water sample became enriched in Ra and isolated from the source (Moore, 2000a). Notice that these approaches are used to estimate where Sshelf and 228Rashelf are the average concentrations of the solute (e. “apparent ages”, defined as the time spent since a water parcel became g., trace metals, nutrients, dissolved organic carbon) of interest and 228 isolated from the Ra source. “Apparent ages” are thus not equivalent Ra over the shelf water column (<200 m). Because the shelf water is with “flushing times”, although they are often used indistinctly in SGD exchanging with the open ocean via mixing, the shelf solute and 228Ra literature (Monsen et al., 2002; Moore et al., 2006). However, evalua­ concentrations must be corrected for their concentrations in the open tions of water apparent ages and flushing times might yield similar time ocean (Socean and 228Raocean). estimates if the system is under steady-state and the comparison of This technique only requires paired measurements of 228Ra and the flushing times and apparent ages is made across an entire study area solute of interest along a shelf-ocean transect. Given that lateral inputs (Tomasky-Holmes et al., 2013). of solutes to the ocean have been shown to be important on a global Average water apparent ages for a system (tw) can be estimated as basis, for example as is the case with dissolved iron (Tagliabue et al., follows, depending on whether the Ra inputs occur at the shoreline 2014), this method is particularly valuable for oceanographic field (Mummy model; Eq. 17) or continuously (Continuous input model; Eq. studies where solute mass balance budgets are required. However, the 18): application of this method requires validating the assumptions of ( ) negligible advection and steady-state conditions on the scale of the tw = ln(ARIn ) − ln ARSys (17) tracer used, which are not valid for all the systems. λs − λl ARIn − ARSys 9. Conclusions and the future stage (>2020) tw = (18) ARSys λs − ARIn λl 9.1. Guidelines to conduct a SGD study where ARIn and ARSys are the activity ratios of the shorter-lived Ra isotope to the longer-lived one in the source (considering all the po­ This article reviews the application of Ra isotopes as tracers of SGD- tential sources) and in the system, respectively, and λs and λl are the derived inputs of water and solutes to the coastal ocean. In this final decay constants of the shorter-lived and longer-lived Ra isotope, chapter, we provide a step-by-step protocol that should serve as respectively. The term λl can be neglected when a long-lived Ra isotope simplified guidelines to perform a SGD study using Ra isotopes. This is used (226Ra and/or 228Ra). Both models assume that i) Ra activities protocol is based on the 7 steps illustrated in Fig. 12. and ARs are highest in the source (ARin > ARSys), ii) the ARin is constant, Step 1 – Definition of the objective of the study: SGD studies often iii) the only Ra losses are due to mixing and radioactive decay, and iv) focus on determining the significance of SGD in the water cycle or on its the open ocean contains negligible activities of the Ra isotopes used implications for coastal biogeochemical cycles. This objective de­ (Charette et al., 2001; Knee et al., 2011; Moore, 2006b; Moore, 2000a). termines the spatial scale of the study (e.g., nearshore vs shelf scale; The activity ratio of any pair of Ra isotopes can be used to estimate water evaluate spatial variability vs produce integrated estimates; spatial apparent ages, provided that the half-life of the shorter-lived Ra isotope resolution), as well as the temporal scale (e.g., snap-shot observations vs is appropriate for the water mixing time-scales expected in the study continuous observations; base conditions vs episodic events; temporal area (Moore, 2015). Typical water apparent ages estimated using 223Ra resolution). The method(s) used to quantify SGD (e.g., water budgets, and 224Ra as the shorter-lived isotope range from ~1 to ~50 days (e.g., groundwater flow models, seepage meter or natural tracers such as Dulaiova et al., 2009; Hancock et al., 2006; Krall et al., 2017; Moore radium, radon, salinity or heat) should be selected according to the et al., 2006; Rengarajan and Sarma, 2015; Sanial et al., 2015). The ac­ specific objective of the study and the characteristics of the system. In curacy of these AR-based apparent water ages will depend on the case of using Ra isotopes, a clear definition of the objective will facilitate analytical uncertainties associated with the isotopes used (e.g., un­ progressing towards the following steps. certainties of 50–100% when water apparent ages are shorter than 3–5 Step 2 – Characterization of the study site: Identifying the hy­ days (Knee et al., 2011). It should be noted that if water masses with drological (e.g., magnitude of surface water inputs, groundwater table different Ra content mix, the relative fractions of each water should be elevation, system geology) and oceanographic characteristics (e.g., cir­ known for reliable apparent age estimates, otherwise the apparent age culation drivers, system timescales, water depths and stratification) will tentatively be biased towards the younger water mass (Delhez et al., characteristics, the potential Ra sources and sinks, as well as the po­ 2003; Hougham and Moran, 2007). tential SGD pathways discharging to the study site (see Fig. 3). This characterization usually involves a combination of literature review, potentially preliminary samplings and the use of complementary tech­ 8.4. Ra-228 as tracer of offshore solute fluxes niques (e.g., salinity profiles, electrical resistivity tomography (ERT), thermal infrared (TIR) images, Rn surveys, water level measurements) Solute fluxes from the coastal ocean derived from SGD, shelves and from different disciplines (e.g., hydrogeology, geophysics, other continental sources have the potential to contribute to biogeo­ oceanography). chemical cycles of the open ocean via shelf-ocean exchange processes. Step 3 – Construction of a conceptual model: A conceptual model However, while solute inputs to the coastal zone are relatively should consider the main SGD pathways, the predominant sources and straightforward to obtain (e.g., Homoky et al., 2016), calculating their sinks of Ra isotopes and the characteristics of the study site (e.g., water net input to the ocean has been a greater challenge owing to solute residence time). The conceptual model is fundamental to selecting the removal or addition processes that take place in estuaries or over the appropriate Ra isotope(s) to be used, and to determining the quantifi­ shelf. To address this shortcoming, Charette et al. (2016) proposed the cation approach, assumptions made and overall sampling approach. It is use of 228Ra as a shelf solute flux gauge, which takes advantage of the important that researchers clearly identify the assumptions of the global shelf 228Ra flux model developed by Kwon et al. (2014). The approach they are applying and validate them. This step is crucial to method is most easily applied where shelf-ocean exchange is primarily characterize the limitations of the method used, to constrain the 18 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Fig. 12. Recommended step-by-step protocol to conduct SGD studies by using Ra isotopes based on the current state of the art. uncertainties of the final estimates and, thus, to produce reliable and biogeochemical cycles (see Section 7.6). justifiable SGD estimates (see section 7.1). Step 4 – Selection of the Ra isotope(s): Ra isotopes applied in the study should be chosen according to the target process or pathway, the 9.2. Knowledge gaps and research needs enrichment in Ra of discharging groundwaters, the potential sources and sinks in the study site and the residence time of the isotope in the study Despite the advances in the use of Ra isotopes as SGD tracer over the area (see Sections 7.3 and 7.4). last decades reviewed in this article, several major research gaps remain Step 5 – Sampling Ra in all the compartments: Samples for Ra open. We believe that the following issues should be addressed in future isotopes should be collected both in seawater and in all the potential studies to improve the use of Ra isotopes as a tracer in SGD studies. endmembers (SGD from different pathways, open ocean, surface water, Discriminating SGD pathways surface discharges, etc.). The collection of sediment samples or cores As highlighted in this review, identifying the SGD pathways or might also be required to evaluate inputs of Ra from sediments. Aside mechanisms that are most significant in contributing Ra isotopes at a from Ra isotopes, concurrent samples for other parameters (e.g., specific study site is crucial to obtaining realistic SGD estimates. There is salinity, water composition, stable isotopes (δ2H, δ18O) of water, con­ thus a need to obtain a conceptual understanding of the systems prior to centrations of nutrients, metals, contaminants) might also be collected the application of Ra isotopes to trace SGD. The combination of different depending on the objectives of the study. Ra isotopes can be instrumental to quantify not only the total magnitude Step 6 – Quantification of SGD using Ra: The quantification of SGD, but also the respective water flows supplied by different path­ approach (mass balance, endmember mixing model or export offshore) ways. Their discrimination will allow obtaining accurate estimations of must be chosen according to the characteristics of the study site and the solute fluxes supplied by SGD, which will decisively contribute to­ considering the validity of model assumptions (see section 6.1). The wards a better understanding of the role SGD plays in coastal biogeo­ quantification of SGD requires constraining the sources and sinks of Ra chemical cycles. isotopes in the study area (step 6.1) and characterizing the appropriate Multi-method studies Ra concentration in the SGD endmembers (step 6.2). In complex set­ Comparisons of Ra-based SGD estimates with independent methods tings, the contribution from the different SGD pathways needs to be (e.g., other tracers, seepage meters, hydrological modelling) are a key accurately accounted for to provide reliable SGD estimates. We note that step towards the validation and refinement of Ra-derived estimates. this may not be possible in some study sites, and so the use of other However, estimates derived from different methods are not always tracers should be considered. Ideally, the estimates derived from Ra directly comparable because different approaches often capture isotopes should be compared with other methods (e.g., seepage meters, different components of SGD or integrate over unique spatio-temporal other tracers, hydrogeological models) to validate and better constrain scales. Using Ra isotopes to discriminate SGD pathways will thus facil­ the results obtained. itate multi-method comparisons and, at the same time, multi-method Step 6.1 – Constraining Ra sources and sinks: The accuracy of studies can contribute to distinguishing different SGD components. SGD estimates strongly rely on the appropriate quantification of the Spatial and temporal Ra variability within the subterranean relevant sources and sinks of Ra at the study site. The significance of estuary sources and sinks (and thus the accuracy with which these parameters There are still few studies constraining the spatial and temporal need to be determined) depend on the characteristics of the study site variability of Ra isotopes in the subsurface and Ra behaviour in the (Fig. 8 and 9) and the Ra isotope used (see 7.3). subterranean estuary is not properly understood. Field measurements Step 6.2 – Determination of the SGD endmember(s): The deter­ should also be combined with reactive transport models to understand mination of the Ra concentration in the SGD endmember(s) is often the the spatio-temporal variability of Ra isotopes within the subterranean major source of uncertainty of SGD estimates. In coastal system with a estuary and the role of physical mechanisms driving groundwater. dominant SGD pathway, the Ra endmember concentration should be Development of new analytical techniques for the measurement representative of this pathway. However, in more complex settings, the of Ra isotopes respective Ra endmembers for each of the pathways as well as its pro­ Most of the methods applied to measure Ra isotopes require pre- portion of SGD contribution should be determined when possible (see concentration of Ra from large sample volumes (usually 10–200 L) Sections 6.2 and 7.5). Ra-based SGD assessments should focus their and measurements via radiometric techniques such as the RaDeCC sys­ effort on accurately determining these endmembers. tem, gamma spectrometry, ICP-MS and TIMS systems. However, there is Step 7 – Quantification of SGD-driven solute fluxes: A represen­ still the need for an accurate and easy methodology to measure the four tative concentration of solutes or solute/Ra ratios in the SGD end­ Ra isotopes in small volumes of water (<1 L), allowing for the routine member/s needs to be determined to estimate SGD-driven solute fluxes. analysis of a large number of samples, and for determining Ra concen­ Importantly, the solute composition of different SGD pathways may be trations in compartments where only small volumes can be sampled (e. considerably different and specific endmembers should be determined g., porewaters, deep ocean). These developments would facilitate the for the different components. This is especially important for studies that improvement of the spatial and temporal resolution of Ra samples aim to quantify SGD-driven solute fluxes (nutrients, metals, contami­ collected both in the subterranean estuary and the ocean and would nants, etc) or to evaluate the significance of SGD in coastal allow, for instance, to assess Ra concentration changes over time, obtaining detailed Ra profiles in the water column or high-resolution Ra 19 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 distributions in the subterranean estuary. These methodological im­ Canals, M., 2019. Remobilization of dissolved metals from a coastal mine tailing deposit driven by groundwater discharge and porewater exchange. Sci. Total provements are thus crucial to move towards a better understanding of Environ. 688, 1359–1372. https://doi.org/10.1016/j.scitotenv.2019.06.224. the different SGD pathways and their spatio-temporal scales. Alorda-Kleinglass, A., Ruiz-Mallén, I., Diego-Feliu, M., Rodellas, V., Bruach-Menchén, J. Uncertainties of Ra-derived SGD estimates M., Garcia-Orellana, J., 2021. The social implications of Submarine Groundwater SGD fluxes obtained from Ra isotopes have large uncertainties that Discharge from an Ecosystem Services perspective: A systematic review. Earth- Science Rev. 103742. are frequently overlooked. Propagated uncertainties associated with Amir, N., Kafri, U., Herut, B., Shalev, E., 2013. Numerical Simulation of Submarine estimates should always and clearly be reported in SGD studies to Groundwater Flow in the Coastal Aquifer at the Palmahim Area, the Mediterranean facilitate understanding of the precision of tracer approaches. Impor­ Coast of Israel. Water Resour. Manag. 27, 4005–4020. https://doi.org/10.1007/ s11269-013-0392-2. tantly, an accurate assessment of uncertainties should not only consider Andrisoa, A., Stieglitz, T.C., Rodellas, V., Raimbault, P., 2019. Primary production in the uncertainties linked to individual parameters (e.g., analytical errors coastal lagoons supported by groundwater discharge and porewater fluxes inferred of Ra measurements, the standard deviation of Ra averages), but also from nitrogen and carbon isotope signatures. Marine Chemistry 210, 48–60. https:// doi.org/10.1016/j.marchem.2019.03.003. those errors linked to the conceptualization of the system (e.g., Annett, A.L., Henley, S.F., Van Beek, P., Souhaut, M., Ganeshram, R., Venables, H.J., assumption of steady state, selection of the endmember, assumptions Meredith, M.P., Geibert, W., van Beek, P., Souhaut, M., Ganeshram, R., Venables, H. linked to mixing loss assessments). J., Meredith, M.P., Geibert, W., 2013. Use of radium isotopes to estimate mixing rates and trace sediment inputs to surface waters in northern Marguerite Bay, Towards more robust Ra-derived SGD estimates Antarctic Peninsula. Antarct. Sci. 25, 445–456. https://doi.org/10.1017/ All studies using Ra isotopes to quantify SGD processes are based on S0954102012000892. numerous assumptions that should be validated in order to produce Anwar, N., Robinson, C., Barry, D., 2014. Influence of tides and waves on the fate of nutrients in a nearshore aquifer: Numerical simulations. Adv. Water Resour. 73, accurate SGD estimates. In other words, most of the investigations using 203–213. https://doi.org/10.1016/j.advwatres.2014.08.015. Ra isotopes produce SGD estimates, but the estimated fluxes are only Baudron, P., Cockenpot, S., Lopez-Castejon, F., Radakovitch, O., Gilabert, J., Mayer, A., meaningful and reliable if the assumptions and uncertainties of the Garcia-Arostegui, J.L., Martinez-Vicente, D., Leduc, C., Claude, C., 2015. Combining model are properly understood, acknowledged, quantified, and radon, short-lived radium isotopes and hydrodynamic modeling to assess submarine groundwater discharge from an anthropized semiarid watershed to a Mediterranean accounted for. Authors working on SGD-related investigations should lagoon (Mar Menor, SE Spain). J. Hydrol. 525, 55–71. https://doi.org/10.1016/j. always be aware of the limitations of the tracers and approaches used. In jhydrol.2015.03.015. some cases, traditional approaches (or the tracer itself) might not be Beck, A.J., Cochran, M.A., 2013. Controls on solid-solution partitioning of radium in saturated marine sands. Mar. Chem. 156, 38–48. https://doi.org/10.1016/j. appropriate for the system studied or the objective of the investigation, marchem.2013.01.008. requiring the development of more complex models or the use of Beck, A.J., Rapaglia, J.P., Cochran, J.K., Bokuniewicz, H.J., 2007. Radium mass-balance alternative/complementary methods. in Jamaica Bay, NY: Evidence for a substantial flux of submarine groundwater. Mar. Chem. 106, 419–441. https://doi.org/10.1016/j.marchem.2007.03.008. Beck, A.J., Rapaglia, J.P., Cochran, J.K., Bokuniewicz, H.J., Yang, S., 2008. Submarine Declaration of Competing Interest groundwater discharge to Great South Bay, NY, estimated using Ra isotopes. Mar. Chem. 109, 279–291. https://doi.org/10.1016/j.marchem.2007.07.011. Beck, A.J., Charette, M.A., Cochran, J.K., Gonneea, M.E., Peucker-Ehrenbrink, B., 2013. None Dissolved strontium in the subterranean estuary – Implications for the marine strontium isotope budget. Geochim. Cosmochim. Acta 117, 33–52. Bejannin, S., Tamborski, J.J., van Beek, P., Souhaut, M., Stieglitz, T., Radakovitch, O., Acknowledgements Claude, C., Conan, P., Pujo-Pay, M., Crispi, O., Le Roy, E., Estournel, C., 2020. Nutrient Fluxes Associated With Submarine Groundwater Discharge From Karstic The authors would like to thank T. Horscroft for the invitation to Coastal Aquifers (Côte Bleue, French Mediterranean Coastline). Front. Environ. Sci. 7, 1–20. https://doi.org/10.3389/fenvs.2019.00205. write this review and W.S. Moore, S. Lamontagne and three anonymous Beneš, P., Strejc, P., Lukavec, Z., 1984. Interaction of radium with freshwater sediments reviewers for their constructive comments on the manuscript. J.Garcia- and their mineral components. I. - Ferric hydroxide and quartz. J. Radioanal. Nucl. Orellana acknowledges the financial support of the Spanish Ministry of Chem. Artic. 82, 275–285. https://doi.org/10.1007/BF02037050. Beneš, P., Borovec, Z., Strejc, P., 1985. Interaction of radium with freshwater sediments Science, Innovation and Universities, through the “Maria de Maeztu” and their mineral components – II. Kaolinite and montmorillonite. J. Radioanal. programme for Units of Excellence (CEX2019-000940-M), the Gen­ Nucl. Chem. Artic. 89, 339–351. https://doi.org/10.1007/BF02040598. eralitat de Catalunya (MERS; 2017 SGR – 1588) and the project OPAL Bethke, C.M., Johnson, T.M., 2002. Paradox of groundwater age. Geology 30, 107–110. https://doi.org/10.1130/0091-7613. (PID2019-110311RB-C21). V. Rodellas acknowledges financial support Blanchard, R.L., Oakes, D., 1965. Relationships between uranium and radium in coastal from the Beatriu de Pinós postdoctoral program of the Generalitat de marine shells and their environment. J. Geophys. Res. 70, 2911–2921. https://doi. Catalunya (2017-BP-00334 and 2019-BP-00241). M. Charette received org/10.1029/jz070i012p02911. Boehm, A.B., Shellenbarger, G.G., Paytan, A., 2004. Groundwater Discharge: Potential support from the U.S. National Science Foundation (OCE-1736277). J. Association with Fecal Indicator Bacteria in the Surf Zone. Environ. Sci. Technol. 38, Scholten acknowledges the support through the SEAMOUNT BONUS 3558–3566. project (art. 185), which is funded jointly by the EU and the Federal Boehm, A.B., Paytan, A., Shellenbarger, G.G., Davis, K.A., 2006. Composition and flux of Ministry of Education and Research of Germany (BMBF, grant no. groundwater from a California beach aquifer: implications for nutrient supply to the surf zone. Cont. Shelf Res. 26, 269–282. https://doi.org/10.1016/j.csr.2005.11.008. 03F0771B). P. van Beek and T. Stieglitz acknowledge support from the Bokuniewicz, H.J., 1980. Groundwater seepage into Great South Bay, New York. Estuar. French ANR project MED-SGD (ANR-15-01CE-0004) and chair @RAc­ Coast. Shelf Sci. 10, 437–444. https://doi.org/10.1016/S0302-3524(80)80122-8. tion MED-LOC (ANR-14-ACHN-0007-01). A. Alorda-Kleinglass ac­ Bokuniewicz, H.J., 1992. Analytical descriptions of subaqueous groundwater seepage. Estuaries 15, 458. https://doi.org/10.2307/1352390. knowledges financial support from ICTA “Unit of Excellence” (MinECo, Bokuniewicz, H., Cochran, J.K., Garcia-Orellana, J., Rodellas, V., Daniel, J.W., MDM2015-0552-17-1) and PhD fellowship, BES-2017-080740. H. Heilbrun, C., 2015. Intertidal percolation through beach sands as a source of 224, Michael acknowledges support from the U.S. National Science Founda­ 223 Ra to Long Island Sound, New York, and Connecticut, United States. J. Mar. Res. 73, 123–140. tion (EAR-1759879). M. Diego-Feliu acknowledges the financial support Bollinger, M.S., Moore, W.S., 1984. Radium fluxes from a salt marsh. Nature 309, 444. from the FI-2017 fellowships of the Generalitat de Catalunya (2017-FIB- Bollinger, M.S., Moore, W.S., 1993. Evaluation of salt marsh hydrology using radium as a 00365). Figs. 3, 4, 7 and 12 were designed by Gemma Solà (www.gemm tracer. Geochim. Cosmochim. Acta 57, 2203–2212. Boudreau, B.P., 1997a. Diagenetic Models and their Implementation. Springer Verlag, asola.com). Berlin, Heidelberg. Boudreau, B.P., 1997b. Diagenetic Models and Their Implementation: Modelling References Transport and Reactions in Aquatic Sediments. Springer, Velag, Berlin, Heidelberg, NY. Broecker, W.S., Li, Y.-H., John, C., 1967. Radium-226 and Radon-222 : Concentration in Adolf, J.E., Burns, J., Walker, J.K., Gamiao, S., 2019. Near shore distributions of Atlantic and Pacific Oceans. Science (80-.) 158, 1307–1311. https://doi.org/ phytoplankton and bacteria in relation to submarine groundwater discharge-fed 10.1126/science.158.3806.1307. fishponds, Kona coast, Hawai‘i, USA. Estuar. Coast. Shelf Sci. 219, 341–353. https:// Burnett, W.C., Cowart, J.B., Deetae, S., 1990. Radium in the Suwannee River and Estuary doi.org/10.1016/j.ecss.2019.01.021. Spring and river input to the Gulf of Mexico. Biogeochemistry 10, 237–255. Alorda-Kleinglass, A., Garcia-Orellana, J., Rodellas, V., Cerdà-Domènech, M., Tovar- Sánchez, A., Diego-Feliu, M., Trezzi, G., Sánchez-Quilez, D., Sanchez-Vidal, A., 20 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Burnett, W., Taniguchi, M., Oberdorfer, J., 2001. Measurement and significance of the Chen, C., Liu, H., Beardsley, R.C., 2003. An unstructured grid, finite-volume, three- direct discharge of groundwater into the coastal zone. J. Sea Res. 46, 109–116. dimensional, primitive equations ocean model: Application to coastal ocean and Burnett, B., Chanton, J., Christoff, J., Kontar, E., Krupa, S., Lambert, M., Moore, W., estuaries. J. Atmos. Ocean. Technol. 20, 159–186. https://doi.org/10.1175/1520- O’Rourke, D., Paulsen, R., Smith, C., Smith, L., Taniguchi, M., 2002. Assessing 0426(2003)020<0159:AUGFVT>2.0.CO;2. methodologies for measuring groundwater discharge to the Ocean. Eos (Washington. Cho, H.-M., Kim, G., 2016. Determining groundwater Ra endmember values for the DC). 83, 1–5. https://doi.org/10.1029/2002EO000069. estimation of the magnitude of submarine groundwater discharge using Ra isotope Burnett, W.C., Bokuniewicz, H., Huettel, M., Moore, W., Taniguchi, M., 2003. tracers. Geophys. Res. Lett. 1–7 https://doi.org/10.1002/2016GL068805. Groundwater and pore water inpunts to the coastal zone. Biogeochemistry 66, 3–33. Cho, H.-M., Kim, G., Kwon, E.Y., Moosdorf, N., Garcia-Orellana, J., Santos, I.R., 2018. Burnett, W.C., Aggarwal, P.K., Aureli, A., Bokuniewicz, H., Cable, J.E., Charette, M.A., Radium tracing nutrient inputs through submarine groundwater discharge in the Kontar, E., Krupa, S., Kulkarni, K.M., Loveless, A., Moore, W.S., Oberdorfer, J.A., global ocean. Sci. Rep. 8, 2439. https://doi.org/10.1038/s41598-018-20806-2. Oliveira, J., Ozyurt, N., Povinec, P., Privitera, A.M.G., Rajar, R., Ramassur, R.T., Church, T.M., 1996. An underground route for the water cycle. Nature 380, 579. Scholten, J., Stieglitz, T., Taniguchi, M., Turner, J.V., 2006. Quantifying submarine Cochran, J.K., Krishnaswami, S., 1980. Radium, thorium, uranium, and Pb210 in deep- groundwater discharge in the coastal zone via multiple methods. Sci. Total Environ. sea sediments and sediment pore waters from the North Equatorial Pacific. Am. J. 367, 498–543. https://doi.org/10.1016/j.scitotenv.2006.05.009. Sci. https://doi.org/10.2475/ajs.280.9.849. Burt, W.J., Thomas, H., Auclair, J.P., 2013a. Short-lived radium isotopes on the Scotian Colbert, S.L., Hammond, D.E., 2007. Temporal and spatial variability of radium in the Shelf: Unique distribution and tracers of cross-shelf CO2 and nutrient transport. Mar. coastal ocean and its impact on computation of nearshore cross-shelf mixing rates. Chem. 156, 120–129. https://doi.org/10.1016/j.marchem.2013.05.007. Cont. Shelf Res. 27, 1477–1500. https://doi.org/10.1016/j.csr.2007.01.003. Burt, W.J., Thomas, H., Fennel, K., Horne, E., 2013b. Sediment-water column fluxes of Colbert, S.L.S.L., Hammond, D.E.D.E., 2008. Shoreline and seafloor fluxes of water and carbon, oxygen and nutrients in Bedford Basin, Nova Scotia, inferred from 224Ra short-lived Ra isotopes to surface water of San Pedro Bay. CA. Mar. Chem. 108, 1–17. measurements. Biogeosciences 10, 53–66. https://doi.org/10.5194/bg-10-53-2013. https://doi.org/10.1016/j.marchem.2007.09.004. Burt, W.J., Thomas, H., Pätsch, J., Omar, A.M., Schrum, C., Daewel, U., Brenner, H., Cook, P.G., Herczeg, A.L., 2012. Environmental Tracers in Subsurface Hydrology. Baar, H.J.W., 2014. Radium isotopes as a tracer of sediment-water column exchange Springer Science, New York. in the North Sea. Glob. Biogeochem. Cycles 786–804. https://doi.org/10.1002/ Cook, P.G., Rodellas, V., Stieglitz, T.C., 2018. Quantifying surface water, porewater and 2014GB004825.Received. groundwater interactions using tracers: tracer fluxes, water fluxes and endmember Cai, P., Shi, X., Moore, W.S., Dai, M., 2012. Measurement of Ra-224:Th-228 concentrations. Water Resour. Res. 1–14. https://doi.org/10.1002/2017WR021780. disequilibrium in coastal sediments using a delayed coincidence counter. Mar. Chem. Cooper, H.H., 1959. A hypothesis concerning the dynamic balance of fresh water and salt 138, 1–6. https://doi.org/10.1016/j.marchem.2012.05.004. water in a coastal aquifer. J. Geophys. Res. 64, 461–467. https://doi.org/10.1029/ Cai, P., Shi, X., Moore, W.S., Peng, S., Wang, G., Dai, M., 2014. 224Ra:228Th JZ064i004p00461. disequilibrium in coastal sediments: Implications for solute transfer across the Copenhaver, S.A., Krishnaswami, S., Turekian, K.K., Shaw, H., 1992. 238U and 232Th sediment-water interface. Geochim. Cosmochim. Acta 125, 68–84. https://doi.org/ series nuclides in groundwater from the J-13 well at the Nevada Test Site: 10.1016/j.gca.2013.09.029. implications for ion retardation. Geophys. Res. Lett. 19, 1383–1386. https://doi.org/ Cai, P., Shi, X., Hong, Q., Li, Q., Liu, L., Guo, X., Dai, M., 2015. Using 224Ra/228Th 10.1029/92GL01437. disequilibrium to quantify benthic fluxes of dissolved inorganic carbon and nutrients Copenhaver, S.A., Krishnaswami, S., Turekian, K.K., Epler, N., Cochran, J.K., 1993. into the Pearl River Estuary. Geochim. Cosmochim. Acta 170, 188–203. https://doi. Retardation of 238U and 232Th decay chain radionuclides in Long Island and org/10.1016/j.gca.2015.08.015. Connecticut aquifers. Geochim. Cosmochim. Acta 57, 597–603. https://doi.org/ Cai, P., Wei, L., Geibert, W., Koehler, D., Ye, Y., Liu, W., Shi, X., 2020. Carbon and 10.1016/0016-7037(93)90370-C. nutrient export from intertidal sand systems elucidated by 224Ra/228Th disequilibria. Danielescu, S., MacQuarrie, K.T.B., Faux, R.N., 2009. The integration of thermal infrared Geochim. Cosmochim. Acta 274, 302–316. https://doi.org/10.1016/j. imaging, discharge measurements and numerical simulation to quantify the relative gca.2020.02.007. contributions of freshwater inflows to small estuaries in Atlantic Canada. Hydrol. Capone, D.G., Bautista, M.F., 1985. A groundwater source of nitrate in nearshore marine Process. 23, 2847–2859. https://doi.org/10.1002/hyp.7383. sediments. Nature 318, 806–809. https://doi.org/10.1038/316507a0. Davidson, M.R., Dickson, B.L., 1986. A porous flow model for steady state transport of Cerdà-Domènech, M., Rodellas, V., Folch, A., Garcia-Orellana, J., 2017. Constraining the radium in groundwater. Water Resour. Res. 22, 34–44. https://doi.org/10.1029/ temporal variations of Ra isotopes and Rn in the groundwater end-member: WR022i001p00034. Implications for derived SGD estimates. Sci. Total Environ. 595, 849–857. https:// Delhez, É.J.M., Deleersnijder, É., Mouchet, A., Beckers, J.M., 2003. A note on the age of doi.org/10.1016/j.scitotenv.2017.03.005. radioactive tracers. J. Mar. Syst. 38, 277–286. https://doi.org/10.1016/S0924-7963 Chabaux, F., Riotte, J., Dequincey, O., 2003. U-Th-Ra fractionation during weathering (02)00245-2. and river transport. Uranium-series Geochemistry 52, 533–576. https://doi.org/ Dias, T.H., de Oliveira, J., Sanders, C.J., Carvalho, F., Sanders, L.M., Machado, E.C., 10.2113/0520533. Sá, F., 2016. Radium isotope (223Ra,224Ra,226Ra and228Ra) distribution near Brazil’s Chan, L.H., Drummond, D., Edmond, J.M., Grant, B., 1977. On the barium data from the largest port, Paranaguá Bay. Brazil. Mar. Pollut. Bull. 111, 443–448. https://doi.org/ Atlantic GEOSECS expedition. Deep. Res. 24, 613–649. https://doi.org/10.1016/ 10.1016/j.marpolbul.2016.07.004. 0146-6291(77)90505-7. Diego-Feliu, M., Rodellas, V., Alorda-Kleinglass, A., Tamborski, J., van Beek, P., Charette, M.A., 2007. Hydrologic forcing of submarine groundwater discharge: Insight Heins, L., Bruach-Menchén, J.M., Arnold, R., Garcia-Orellana, J., 2020. Guidelines from a seasonal study of radium isotopes in a groundwater-dominated salt marsh and limits for the quantification of Ra isotopes and related radionuclides with the estuary. Limnol. Oceanogr. 52, 230–239. Radium Delayed Coincidence Counter (RaDeCC). J. Geophys. Res. C Ocean 125. Charette, M.a., Buesseler, K.O., 2004. Submarine groundwater discharge of nutrients and https://doi.org/10.1029/2019JC015544. copper to an urban subestuary of Chesapeake Bay (Elizabeth River). Limnol. Diego-Feliu, M., Rodellas, V., Saaltink, M.W., Alorda-Kleinglass, A., Goyetche, T., Oceanogr. 49, 376–385. https://doi.org/10.4319/lo.2004.49.2.0376. Martínez-Pérez, L., Folch, A., Garcia-Orellana, J., 2021. New perspectives on the use Charette, M.A., Sholkovitz, E.R., 2006. Trace element cycling in a subterranean estuary: of 224Ra/228Ra and 222Rn/226Ra activity ratios in groundwater studies. J. Hydrol. Part 2. Geochemistry of the pore water. Geochim. Cosmochim. Acta 70, 811–826. 596, 126043. https://doi.org/10.1016/j.jhydrol.2021.126043. https://doi.org/10.1016/j.gca.2005.10.019. Dulaiova, H., Burnett, W.C., Chanton, J.P., Moore, W.S., Bokuniewicz, H.J., Charette, M. Charette, M.A., Buesseler, K.O., Andrews, J.R.E., 2001. Utility of radium isotopes for A., Sholkovitz, E., 2006a. Assessment of groundwater discharges into West Neck Bay, evaluating the input and transport of groundwater-derived nitrogen to a Cape Cod New York, via natural tracers. Cont. Shelf Res. 26, 1971–1983. https://doi.org/ estuary. Limnol. Oceanogr. 46, 465–470. 10.1016/j.csr.2006.07.011. Charette, M.A., Splivallo, R., Herbold, C., Bollinger, M.S., Moore, W.S., 2003. Salt marsh Dulaiova, H., Burnett, W.C., Wattayakorn, G., Sojisuporn, P., 2006b. Are groundwater submarine groundwater discharge as traced by radium isotopes. Mar. Chem. 84, inputs into river-dominated areas important? The Chao Phraya River - Gulf of 113–121. https://doi.org/10.1016/j.marchem.2003.07.001. Thailand. Limnol. Oceanogr. 51, 2232–2247. https://doi.org/10.4319/ Charette, M.A., Gonneea, M.E., Morris, P.J., Statham, P., Fones, G., Planquette, H., lo.2006.51.5.2232. Salter, I., Garabato, A.N., 2007. Radium isotopes as tracers of iron sources fueling a Dulaiova, H., Ardelan, M.V., Henderson, P.B., Charette, M.A., 2009. Shelf-derived iron Southern Ocean phytoplankton bloom. Deep Sea Res. Part II Top. Stud. Oceanogr. inputs drive biological productivity in the southern Drake Passage. Glob. 54, 1989–1998. https://doi.org/10.1016/j.dsr2.2007.06.003. Biogeochem. Cycles 23, 1–14. https://doi.org/10.1029/2008GB003406. Charette, M.A., Moore, W.S., Burnett, W.C., 2008. Uranium-and Thorium- Series Duque, C., Knee, K.L., Russoniello, C.J., Sherif, M., Abu Risha, U.A., Sturchio, N.C., Nuclides as Tracers of Submarine Groundwater Discharge CHAPTER-5 Uranium-and Michael, H.A., 2019. Hydrogeological processes and near shore spatial variability of Thorium-Series Nuclides as Tracers of Submarine Groundwater Discharge. Radioact. radium and radon isotopes for the characterization of submarine groundwater Environ. 13, 234–289. https://doi.org/10.1016/S1569-4860(07)00005-8. discharge. J. Hydrol. 579, 124192. https://doi.org/10.1016/j.jhydrol.2019.124192. Charette, M.A., Henderson, P.B., Breier, C.F., Liu, Q., 2013. Submarine groundwater Duque, C., Michael, H.A., Wilson, A.M., 2020. The subterranean estuary: Technical term, discharge in a river-dominated Florida estuary. Mar. Chem. 156, 3–17. https://doi. simple analogy, or source of confusion? Water Resour. Res. 1–7 https://doi.org/ org/10.1016/j.marchem.2013.04.001. 10.1029/2019wr026554. Charette, M.A., Morris, P.J., Henderson, P.B., Moore, W.S., 2015. Radium isotope Dyer, K., 1973. Estuaries: A Physical Introduction. John Wiley, London. distributions during the US GEOTRACES North Atlantic cruises. Mar. Chem. 177, Elsinger, R.J., Moore, W.S., 1980. 226Ra behavior in the Pee Dee River-Winyah Bay 1–12. https://doi.org/10.1016/j.marchem.2015.01.001. estuary. Earth Planet. Sci. Lett. 48, 239–249. https://doi.org/10.1016/0012-821X Charette, M.A., Lam, P.J., Lohan, M.C., Kwon, E.Y., Hatje, V., Jeandel, C., Shiller, A.M., (80)90187-9. Cutter, G., Thomas, A.L., Boyd, P.W., Homoky, W.B., Milne, A., Thomas, H., Elsinger, R.J., Moore, W.S., 1983a. 224Ra,228Ra, and226Ra in Winyah Bay and Delaware Andersson, P.S., Porcelli, D., Tanaka, T., Geibert, W., Dehairs, F., Garcia-Orellana, J., Bay. Earth Planet. Sci. Lett. 64, 430–436. https://doi.org/10.1016/0012-821X(83) 2016. Coastal ocean and shelf-sea biogeochemical cycling of trace elements and 90103-6. isotopes: lessons learned from GEOTRACES. Trans. R. Soc. 374 https://doi.org/ Elsinger, Robert J., Moore, W.S., 1983b. 224Ra, 228Ra, and 226Ra in Winyah Bay and 10.1098/rsta.2016.0076. Delaware Bay. Earth Planet. Sci. Lett. 64, 430–436. 21 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Eriksen, D., Sidhu, R., Ramsøy, T., Strålberg, E., Iden, K.I., Rye, H., Hylland, K., Ruus, A., Hancock, G., Webster, I., Ford, P., Moore, W., 2000. Using Ra isotopes to examine Berntssen, M.H.G., 2009. Radioactivity in produced water from Norwegian oil and transport processes controlling benthic fluxes into a shallow estuarine lagoon. gas installations - Concentrations, bioavailability, and doses to marine biota. Geochim. Cosmochim. Acta 64, 3685–3699. https://doi.org/10.1016/S0016-7037 Radioprotection 44, 869–874. https://doi.org/10.1051/radiopro/20095155. (00)00469-5. Erostate, M., Huneau, F., Garel, E., Ghiotti, S., Vystavna, Y., Garrido, M., Pasqualini, V., Hancock, G.J., Webster, I.T., Stieglitz, T.C., 2006. Horizontal mixing of great Barrier Reef 2020. Groundwater dependent ecosystems in coastal Mediterranean regions: waters: offshore diffusivity determined from radium isotope distribution. Characterization, challenges and management for their protection. Water Res. 172, J. Geophys. Res. Ocean. 111, 1–14. https://doi.org/10.1029/2006JC003608. 115461. Homoky, W.B., Weber, T., Berelson, W.M., Conway, T.M., Henderson, G.M., Van Evans, R.D., Kip, A.F., 1938. The radium and radon content of Pacific Ocean water, life, Hulten, M., Jeandel, C., Severmann, S., Tagliabue, A., 2016. Quantifying trace and sediments. Am. J. Sci. 214, 241–259. element and isotope fluxes at the ocean-sediment boundary: A review. Philos. Trans. Evans, T.B., Wilson, A.M., 2017. Submarine groundwater discharge and solute transport R. Soc. A Math. Phys. Eng. Sci. https://doi.org/10.1098/rsta.2016.0246. under a transgressive barrier island. J. Hydrol. 547, 97–110. https://doi.org/ Hong, Q., Cai, P., Shi, X., Li, Q., Wang, G., 2017. Solute transport into the Jiulong River 10.1016/j.jhydrol.2017.01.028. estuary via pore water exchange and submarine groundwater discharge: New Ferrarin, C., Rapaglia, J., Zaggia, L., Umgiesser, G., Zuppi, G.M., 2008. Coincident insights from 224Ra/228Th disequilibrium. Geochim. Cosmochim. Acta 198, application of a mass balance of radium and a hydrodynamic model for the seasonal 338–359. https://doi.org/10.1016/j.gca.2016.11.002. quantification of groundwater flux into the Venice Lagoon, Italy. Mar. Chem. 112, Hong, Q., Cai, P., Geibert, W., Cao, Z., Stimac, I., Liu, L., Li, Q., 2018. Benthic fluxes of 179–188. https://doi.org/10.1016/j.marchem.2008.08.008. metals into the Pearl River Estuary based on 224Ra/228Th disequilibrium: from Folch, A., Val, L., Luquot, L., Martínez-pérez, L., Bellmunt, F., Le Lay, H., 2020. alkaline earth (Ba) to redox sensitive elements (U, Mn, Fe). Geochim. Cosmochim. Combining fi ber optic DTS, cross-hole ERT and time-lapse induction logging to Acta 237, 223–239. https://doi.org/10.1016/j.gca.2018.06.036. characterize and monitor a coastal aquifer. J. Hydrol. 588, 125050. https://doi.org/ Hougham, A.L., Moran, S.B., 2007. Water mass ages of coastal ponds estimated using 10.1016/j.jhydrol.2020.125050. 223Ra and 224Ra as tracers. Mar. Chem. 105, 194–207. https://doi.org/10.1016/j. Freeze, R.A., Cherry, J.A., 1979. Groundwater. Prentice Hall, Englewood Cliff’s, New marchem.2007.01.013. Jersey. Huh, C.A., Ku, T.L., 1998. A 2-D section of 228Ra and 226Ra in the Northeast Pacific. Garces, E., Basterretxea, G., Tovar-Sanchez, A., 2011. Changes in microbial communities Oceanol. Acta 21, 533–542. https://doi.org/10.1016/S0399-1784(98)80036-4. in response to submarine groundwater input. Mar. Ecol. Prog. Ser. 438, 47–58. Hussain, N., 1995. Supply rates of natural U-Th series radionuclides from aquifer solids https://doi.org/10.3354/meps09311. into groundwater. Geophys. Res. Lett. 22, 1521–1524. https://doi.org/10.1029/ Garcia-Orellana, J., Cochran, J.K., Bokuniewicz, H., Yang, S., Beck, A.J., 2010. Time- 95GL01006. series sampling of 223Ra and 224Ra at the inlet to Great South Bay (New York): A Hwang, D.W., Lee, Y.W., Kim, G., 2005. Large submarine groundwater discharge and strategy for characterizing the dominant terms in the Ra budget of the bay. https:// benthic eutrophication in Bangdu Bay on volcanic Jeju Island, Korea. Limnol. doi.org/10.1016/j.jenvrad.2009.12.005. Oceanogr. 50, 1393–1403. Garcia-Orellana, J., Cochran, J.K., Bokuniewicz, H., Daniel, J.W.R., Rodellas, V., Hwang, D.-W., Lee, I.-S., Choi, M., Kim, T.-H., 2016. Estimating the input of submarine Heilbrun, C., 2014. Evaluation of 224Ra as a tracer for submarine groundwater groundwater discharge (SGD) and SGD-derived nutrients in Geoje Bay, Korea using discharge in Long Island Sound (NY). Geochim. Cosmochim. Acta 141, 314–330. 222Rn-Si mass balance model. Mar. Pollut. Bull. 110, 119–126. https://doi.org/ https://doi.org/10.1016/j.gca.2014.05.009. 10.1016/j.marpolbul.2016.06.073. Garcia-Orellana, J., López-Castillo, E., Casacuberta, N., Rodellas, V., Masqué, P., Ivanovich, M., Harmon, R.S., 1992. Uranium-series Disequilibrium: Applications to Carmona-Catot, G., Vilarrasa, M., García-Berthou, E., 2016. Influence of submarine Earth, Marine, and Environmental Sciences. Clarendon Press. groundwater discharge on 210Po and 210Pb bioaccumulation in fish tissues. Iyengar, M.A.R., Rao, K.N., 1990. Uptake of radium by marine animals. In: The J. Environ. Radioact. 155–156, 46–54. https://doi.org/10.1016/j. Environmental Behaviour of Radium Vol. 1 (Tecnical Reports Series No. 310). IAEA, jenvrad.2016.02.005. Vienna, p. 600. Garcia-Solsona, E., Masqué, P., Garcia-Orellana, J., Rapaglia, J., Beck, A.J., Cochran, J. Jeandel, C., 2016. Overview of the mechanisms that could explain the “Boundary K., Bokuniewicz, H.J., Zaggia, L., Collavini, F., 2008. Estimating Submarine Exchange” at the land-ocean contact. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. Groundwater Discharge Around Isola La Cura, Northern Venice Lagoon (Italy), by 374 https://doi.org/10.1098/rsta.2015.0287. Using the Radium Quartet. https://doi.org/10.1016/j.marchem.2008.02.007. Johannes, R., 1980. The ecological significance of the submarine discharge of Garcia-Solsona, E., Garcia-Orellana, J., Masque, P., Garces, E., Radakovitch, O., groundwater. Mar. Ecol. Prog. Ser. 3, 365–373. https://doi.org/10.3354/ Mayer, A., Estrade, S., Basterretxea, G., 2010a. An assessment of karstic submarine meps003365. groundwater and associated nutrient discharge to a Mediterranean coastal area Joly, J., 1908. On the radium-content of deep-sea sediments. London, Edinburgh, Dublin (Balearic Islands, Spain) using radium isotopes. Biogeochemistry 97, 211–229. Philos. Mag. J. Sci. 16, 190–197. https://doi.org/10.1080/14786440708636499. https://doi.org/10.1007/s10533-009-9368-y. Kadko, D.C., Moore, W.S., 1988. Radiochemical constraints on the crustal residence time Garcia-Solsona, E., Garcia-Orellana, J., Masque, P., Rodellas, V., Mejias, M., of submarine hydrothermal fluids: Endeavour Ridge. Geochim. Cosmochim. Acta 52, Ballesteros, B., Dominguez, J.A., 2010b. Groundwater and nutrient discharge 659–668. https://doi.org/10.1016/0016-7037(88)90328-6. through karstic coastal springs (Castello, Spain). Biogeosciences 7, 2625–2638. Kaufman, A., Trier, R.M., Broecker, W.S., Feely, H.W., 1973. Distribution of 228 Ra in the https://doi.org/10.5194/bg-7-2625-2010. world ocean. J. Geophys. Res. 78, 8827–8848. https://doi.org/10.1029/ Gattacceca, J.C., Mayer, A., Cucco, A., Claude, C., Radakovitch, O., Vallet-Coulomb, C., jc078i036p08827. Hamelin, B., 2011. Submarine groundwater discharge in a subsiding coastal lowland: Kelly, R.P., Moran, S.B., 2002. Seasonal changes in groundwater input to a well-mixed A Ra-226 and Rn-222 investigation in the Southern Venice lagoon. Appl. Geochem. estuary estimated using radium isotopes and implications for coastal nutrient 26, 907–920. https://doi.org/10.1016/j.apgeochem.2011.03.001. budgets. Limnol. Oceanogr. 47, 1796–1807. https://doi.org/10.4319/ Geibert, W., Rodellas, V., Annett, A., van Beek, P., Garcia-Orellana, J., Hsieh, Y.-T., lo.2002.47.6.1796. Masque, P., 2013. 226Ra determination via the rate of 222Rn ingrowth with the Key, R.M., Stallard, R.F., Moore, W., Sarmiento, J.L., 1985. Distribution and flux of Ra- Radium Delayed Coincidence Counter (RaDeCC). Limnol. Oceanogr. Methods 11, 226 and Ra-228 in the Amazon River Estuary. J. Geophys. Res. 90, 6995–7004. 594–603. https://doi.org/10.4319/lom.2013.11.594. Kigoshi, K., 1971. Alpha-recoil thorium-234: dissolution into water and the uranium- George, C., Moore, W.S., White, S.M., Smoak, E., Joye, S.B., Leier, A., Wilson, A.M., 234/uranium-238 disequilibrium in nature. Science (80-.) 173, 47–48. https://doi. 2020. A new mechanism for submarine groundwater discharge from continental org/10.1126/science.173.3991.47. shelves. Water Resour. Res. 56, 1–17. https://doi.org/10.1029/2019WR026866. Kim, G., Ryu, J.W., Yang, H.S., Yun, S.T., 2005. Submarine groundwater discharge (SGD) Giffin, C., Kaufman, A., Broecker, W.S., 1963. Delayed coincidence counter for the assay into the Yellow Sea revealed by Ra-228 and Ra-226 isotopes: implications for global of actinon and thoron. J. Geophys. Res. 68, 1749–1757. silicate fluxes. Earth Planet. Sci. Lett. 237, 156–166. https://doi.org/10.1016/j. Gilfedder, B.S., Frei, S., Hofmann, H., Cartwright, I., 2015. Groundwater discharge to epsl.2005.06.011. wetlands driven by storm and flood events: quantification using continuous Radon- Kim, G., Ryu, J.-W., Hwang, D.-W., 2008. Radium tracing of submarine groundwater 222 and electrical conductivity measurements and dynamic mass-balance modelling. discharge (SGD) and associated nutrient fluxes in a highly-permeable bed coastal Geochim. Cosmochim. Acta 165, 161–177. https://doi.org/10.1016/j. zone, Korea. Mar. Chem. 109, 307–317. https://doi.org/10.1016/j. gca.2015.05.037. marchem.2007.07.002. Glover, R.E., 1959. The pattern of fresh-water flow in a coastal aquifer. J. Geophys. Res. King, J.N., 2012. Synthesis of benthic flux components in the Patos Lagoon coastal zone, 64, 457–459. https://doi.org/10.1029/JZ064i004p00457. Rio Grande do Sul, Brazil. Water Resour. Res. 48, 1–10. https://doi.org/10.1029/ Gonneea, M.E., Charette, M.A., 2014. Hydrologic controls on nutrient cycling in an 2011WR011477. unconfined coastal aquifer. Environ. Sci. Technol. 48, 14178–14185. https://doi. Kiro, Y., Yechieli, Y., Voss, C.I., Starinsky, A., Weinstein, Y., 2012. Modeling radium org/10.1021/es503313t. distribution in coastal aquifers during sea level changes: the Dead Sea case. Gonneea, M.E., Morris, P.J., Dulaiova, H., Charette, M.A., 2008. New perspectives on Geochim. Cosmochim. Acta 88, 237–254. https://doi.org/10.1016/j. radium behavior within a subterranean estuary. Mar. Chem. 109, 250–267. https:// gca.2012.03.022. doi.org/10.1016/j.marchem.2007.12.002. Kiro, Y., Weinstein, Y., Starinsky, A., Yechieli, Y., 2013. Groundwater ages and reaction Gonneea, M.E., Mulligan, A.E., Charette, M.A., 2013. Seasonal cycles in radium and rates during seawater circulation in the Dead Sea aquifer. Geochim. Cosmochim. barium within a subterranean estuary: Implications for groundwater derived Acta 122, 17–35. https://doi.org/10.1016/j.gca.2013.08.005. chemical fluxes to surface waters. Geochim. Cosmochim. Acta 119, 164–177. Knauss, K.G., Ku, T.-L., Moore, W.S., 1978. Radium and thorium isotopes in the surface https://doi.org/10.1016/j.gca.2013.05.034. waters of the East Pacific and coastal Southern California. Earth Planet. Sci. Lett. 39, Gu, H., Moore, W.S., Zhang, L., Du, J., Zhang, J., 2012. Using radium isotopes to estimate 235–249. https://doi.org/10.1016/0012-821X(78)90199-1. the residence time and the contribution of submarine groundwater discharge (SGD) Knee, K.L., Paytan, A., 2011. Submarine Groundwater Discharge: A Source of Nutrients, in the Changjiang effluent plume. East China Sea. Cont. Shelf Res. 35, 95–107. Metals, and Pollutants to the Coastal Ocean. Treatise Estuar. Coast. Sci. 4. https://doi.org/10.1016/j.csr.2012.01.002. 22 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Knee, K.L., Street, J.H., Grossman, E.E., Boehm, A.B., Paytan, A., 2010. Nutrient inputs to Liu, Q., Charette, M.A., Henderson, P.B., McCorkle, D.C., Martin, W., Dai, M., 2014. the coastal ocean from submarine groundwater discharge in a groundwater- Effect of submarine groundwater discharge on the coastal ocean inorganic carbon dominated system: Relation to land use (Kona coast, Hawaii, USA). Limnol. cycle. Limnol. Oceanogr. 59, 1529–1554. https://doi.org/10.4319/ Oceanogr. 55, 1105–1122. https://doi.org/10.4319/lo.2010.55.3.1105. lo.2014.59.5.1529. Knee, K.L., Garcia-Solsona, E., Garcia-Orellana, J., Boehm, A.B., Paytan, A., 2011. Using Liu, J., Su, N., Wang, X., Du, J., 2018. Submarine groundwater discharge and associated radium isotopes to characterize water ages and coastal mixing rates: a sensitivity nutrient fluxes into the Southern Yellow Sea: A case study for semi- enclosed and analysis. Limnol. Oceanogr. 9, 380–395. https://doi.org/10.4319/lom.2011.9.380. oligotrophic seas-implication for green tide bloom. J. Geophys. Res. Ocean. 121, Koch-Larrouy, A., Atmadipoera, A., van Beek, P., Madec, G., Aucan, J., Lyard, F., 3372–3380. https://doi.org/10.1002/2015JC011421.Received. Grelet, J., Souhaut, M., 2015. Estimates of tidal mixing in the Indonesian archipelago Liu, Y., Jiao, J.J., Mao, R., Luo, X., Liang, W., Robinson, C.E., 2019. Spatial from multidisciplinary INDOMIX in-situ data. Deep. Res. Part I Oceanogr. Res. Pap. characteristics reveal the reactive transport of radium isotopes (224 Ra, 223 Ra, 106, 136–153. https://doi.org/10.1016/j.dsr.2015.09.007. & 228 Ra) in an intertidal aquifer. Water Resour. Res. 2019 https://doi.org/ Koczy, F.F., 1958. Natural radium as a tracer in the ocean. In: 2nd U.N. Int. Conf. 10.1029/2019WR024849. WR024849. Peaceful Uses of Atomic Energy, 18. Geneva, pp. 351–357. Luek, J.L.J.L., Beck, A.J.A.J., 2014. Radium budget of the York River estuary (VA, USA) Koczy, F.F., 1963. Natural radio nuclides in the ocean: radioactive tracers in dominated by submarine groundwater discharge with a seasonally variable oceanography. In: International Union of Geodesy and Geophysics Monograph No. groundwater end-member. Mar. Chem. 165, 55–65. https://doi.org/10.1016/j. 20. Tenth Pacific Science Congress. Honolulu, Hawaii, pp. 1–9. marchem.2014.08.001. Kohout, F.A., 1966. Submarine springs: A neglected phenomenon of coastal hydrology. Luijendijk, E., Gleeson, T., Moosdorf, N., 2020. Fresh groundwater discharge insigni fi Hydrology 26, 391–413. cant for the world ’ s oceans but important for coastal ecosystems. Nat. Commun. 11, Kraemer, T.F., 2005. Radium isotopes in Cayuga Lake, New York: Indicators of inflow 1260. https://doi.org/10.1038/s41467-020-15064-8. and mixing processes. Limnol. Oceanogr. 50, 158–168. https://doi.org/10.4319/ Luo, X., Jiao, J.J., Moore, W.S., Cherry, J.A., Wang, Y., Liu, K., 2018. Significant lo.2005.50.1.0158. chemical fluxes from natural terrestrial groundwater rival anthropogenic and fluvial Krall, L., Trezzi, G., Garcia-Orellana, J., Rodellas, V., Mörth, C.-M., Andersson, P., 2017. input in a large-river deltaic estuary. Water Res. 144, 603–615. https://doi.org/ Submarine groundwater discharge at Forsmark, Gulf of Bothnia, provided by Ra 10.1016/j.watres.2018.07.004. isotopes. Mar. Chem. 196, 162–172. https://doi.org/10.1016/j. Ma, Q., Zhang, Y., 2020. Global research trends and hotspots on submarine groundwater marchem.2017.09.003. discharge (SGD): A bibliometric analysis. Int. J. Environ. Res. Public Health 17. Krest, J.M., Harvey, J.W., 2003. Using natural distributions of short-lived radium https://doi.org/10.3390/ijerph17030830. isotopes to quantify groundwater discharge and recharge. Limnol. Oceanogr. 48, Manheim, F.T., 1967. Evidence for submarine discharge of water on the Atlantic 290–298. https://doi.org/10.4319/lo.2003.48.1.0290. continental slope of the southern United States, and suggestions for further search. Krest, J.M., Moore, W.S., Rama, 1999. 226Ra and 228Ra in the mixing zones of the Trans. N. Y. Acad. Sci. 29, 839–853. https://doi.org/10.1111/j.2164-0947.1967. Mississippi and Atchafalaya rivers: Indicators of groundwater input. Mar. Chem. 64, tb02825.x. 129–152. https://doi.org/10.1016/S0304-4203(98)00070-X. Mathieu, G.G., Biscaye, P.E., Lupton, R.A., Hammond, D.E., 1988. System for Krest, J.M., Moore, W.S., Gardner, L.R., Morris, J.T., 2000. Marsh nutrient export measurements of 222Rn at low levels in natural waters. Health Phys. 55. supplied by groundwater discharge: evidence from radium measurements. Glob. McAllister, S.M., Barnett, J.M., Heiss, J.W., Findlay, A.J., MacDonald, D.J., Dow, C.L., Biogeochem. Cycles 14, 167–176. https://doi.org/10.1029/1999GB001197. Luther, G.W., Michael, H.A., Chan, C.S., 2015. Dynamic hydrologic and Krishnaswami, S., Graystein, W.C., Turekian, K.K., Dowd, J.F., 1982. Radium, thorium biogeochemical processes drive microbially enhanced iron and sulfur cycling within and radioactive lead isotopes in groundwaters ’ application to the in situ the intertidal mixing zone of a beach aquifer. Limnol. Oceanogr. 60, 329–345. determination of adsorption-desorption. Water Resour. Res. 18, 1633–1675. https://doi.org/10.1111/lno.10029. Kronfeld, J., Ilani, S., Strull, A., 1991. Radium precipitation and extreme 238U-series McKinley, I.G., Russell Alexander, W., 1993. Assessment of radionuclide retardation: uses disequilibrium along the Dead Sea coast, Israel. Appl. Geochem. 6, 355–361. https:// and abuses of natural analogue studies. J. Contam. Hydrol. 13, 249–259. https://doi. doi.org/10.1016/0883-2927(91)90011-D. org/10.1016/0169-7722(93)90060-6. Ku, T.-L., Luo, S., 2008. Ocean circulation/mixing studies with decay-series isotopes. Meier, H., Zimmerhackl, E., Zeitler, G., Menge, P., 2015. Parameter Studies of Krishnaswami, S., Cochran, J.K.B.T.-R. in the E. (Eds.). In: U-Th Series Nuclides in Radionuclide Sorption in Site-Specific Sediment/Groundwater Systems. Radiochim. Aquatic Systems. Elsevier, pp. 307–344. https://doi.org/10.1016/S1569-4860(07) Acta 66–67. https://doi.org/10.1524/ract.1994.6667.special-issue.277. 00009-5. Meysman, F.J.R., Boudreau, B.P., Middelburg, J.J., 2005. Modeling reactive transport in Kumar, A., Rout, S., Pulhani, V., Kumar, A.V., 2020. A review on distribution coefficient sediments subject to bioturbation and compaction. Geochim. Cosmochim. Acta 69, (Kd) of some selected radionuclides in soil/sediment over the last three decades. 3601–3617. https://doi.org/10.1016/j.gca.2005.01.004. J. Radioanal. Nucl. Chem. 323, 13–26. https://doi.org/10.1007/s10967-019-06930- Michael, H.A., Mulligan, A.E., Harvey, C.F., 2005. Seasonal oscillations in water x. exchange between aquifers and the coastal ocean. Nature 436, 1145–1148. https:// Kwon, E.Y., Kim, G., Primeau, F., Moore, W.S., Cho, H.-M., DeVries, T., Sarmiento, J.L., doi.org/10.1038/nature03935. Charette, M.A., Cho, Y.-K., 2014. Global estimate of submarine groundwater Michael, H.A., Charette, M.A., Harvey, C.F., 2011. Patterns and variability of discharge based on an observationally constrained radium isotope model. Geophys. groundwater flow and radium activity at the coast: A case study from Waquoit Bay. Res. Lett. 1–7 https://doi.org/10.1002/2014GL061574.Received. Massachusetts. Mar. Chem. 127, 100–114. https://doi.org/10.1016/j. Lamontagne, S., Webster, I.T., 2019a. Theoretical assessment of the effect of vertical marchem.2011.08.001. dispersivity on coastal seawater radium distribution. Front. Mar. Sci. 6, 1–10. Molina-Porras, A., Condomines, M., Legeay, P.L., Bailly-Comte, V., Seidel, J.L., 2020. https://doi.org/10.3389/fmars.2019.00357. Radium Isotopes as a Tracer of Water Sources and Mixing in the Vidourle Stream Lamontagne, S., Webster, I.T., 2019b. Cross-shelf transport of submarine groundwater (South of France). Aquat. Geochem. 26, 119–136. https://doi.org/10.1007/s10498- discharge tracers: a sensitivity analysis. J. Geophys. Res. Ocean. 124, 453–469. 020-09371-1. https://doi.org/10.1029/2018JC014473. Monsen, N.E., Cloern, J.E., Lucas, L.V., Monismith, S.G., 2002. A comment on the use of Lecher, A.L., Fisher, A.T., Paytan, A., 2016. Submarine groundwater discharge in flushing time, residence time, and age as transport time scales. Limnol. Oceanogr. Northern Monterey Bay, California: evaluation by mixing and mass balance models. 47, 1545–1553. https://doi.org/10.4319/lo.2002.47.5.1545. Mar. Chem. 179, 44–55. https://doi.org/10.1016/j.marchem.2016.01.001. Montiel, D., Dimova, N., Andreo, B., Prieto, J., García-Orellana, J., Rodellas, V., 2018. Lee, C.M., Jiao, J.J., Luo, X., Moore, W.S., 2012. Estimation of submarine groundwater Assessing submarine groundwater discharge (SGD) and nitrate fluxes in highly discharge and associated nutrient fluxes in Tolo Harbour, Hong Kong. Sci. Total heterogeneous coastal karst aquifers: Challenges and solutions. J. Hydrol. 557, Environ. 433, 427–433. https://doi.org/10.1016/j.scitotenv.2012.06.073. 222–242. https://doi.org/10.1016/j.jhydrol.2017.12.036. Leibundgut, C., Seibert, J., 2011. 2.09 - Tracer Hydrology. in: Wilderer, P.B.T.-T. on W.S. Moore, Willard S., 1969a. Oceanic concentrations of 228Radium. Earth Planet. Sci. Lett. (Ed.),. Elsevier, Oxford, pp. 215–236. https://doi.org/10.1016/B978-0-444-53199- 6, 437–446. https://doi.org/10.1016/0012-821x(69)90113-7. 5.00036-1. Moore, Willard S., 1969b. Measurement of 228Ra and 228Th in Sea Water. J. Geophys. Lerner, D.N., Harris, B., 2009. The relationship between land use and groundwater Res. 74, 694–704. resources and quality. Land Use Policy 26. https://doi.org/10.1016/j. Moore, W.S., 1972. Radium-228: Application to thermocline mixing studies. Earth landusepol.2009.09.005. S265–S273. Planet. Sci. Lett. 16, 421–422. Levy, D.M., Moore, W.S., 1985. Ra in continental shelf waters. Earth Planet. Sci. Lett. 73, Moore, W.S., 1976. Sampling 228Ra in the deep ocean. Deep. Res. 23, 647–651. 226–230. Moore, W.S., 1981. Radium isotopes in the Chesapeake Bay. Estuar. Coast. Shelf Sci. 12, Li, Y.-H., Chan, L.-H., 1979. Desorption of Ba and Ra-226 from river-borne sediments in 713–723. https://doi.org/10.1016/S0302-3524(81)80067-9. the Hudson Estuary. Earth Planet. Sci. Lett. 43, 343–350. Moore, W.S., 1987. Radium 228 in the South Atlantic Bight. J. Geophys. Res. Ocean. 92, Li, Y.-H., Mathieu, G., Biscaye, P., Simpson, H.J., 1977. The flux of Ra-226 from estuarine 5177–5190. https://doi.org/10.1029/JC092iC05p05177. and continental shelf sediments. Earth Planet. Sci. Lett. 37, 241–273. Moore, W.S., 1996. Large groundwater inputs to coastal waters revealed by 226Ra Li, X., Ke, T., Wang, Y., Zhou, T., Li, D., Tong, F., Wen, J., 2020. Hydraulic conductivity enrichments. Nature 380, 612–614. behaviors of Karst Aquifer with conduit-fissure geomaterials. Front. Earth Sci. 8, Moore, W.S., 1997. High fluxes of radium and barium from the mouth of the Ganges- 1–10. https://doi.org/10.3389/feart.2020.00030. Brahmaputra River during low river discharge suggest a large groundwater source. Liao, F., Wang, G., Yi, L., Shi, Z., Cheng, G., Kong, Q., Mu, W., Guo, L., Cheng, K., Earth Planet. Sci. Lett. 150, 141–150. https://doi.org/10.1016/S0012-821X(97) Dong, N., Liu, C., 2020. Applying radium isotopes to estimate groundwater discharge 00083-6. into Poyang Lake, the largest freshwater lake in China. J. Hydrol. 585, 124782. Moore, W.S., 1999. The subterranean estuary: a reaction zone of groundwater and https://doi.org/10.1016/j.jhydrol.2020.124782. seawater. Mar. Chem. 65, 111–125. Lin, L., Liu, Z., 2019. Partial residence times: determining residence time composition in Moore, W.S., 2000a. Ages of continental shelf waters determined from 223Ra and 224Ra. different subregions. Ocean Dyn. 69, 1023–1036. https://doi.org/10.1007/s10236- J. Geophys. Res. 22117–22122. 019-01298-8. 23 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 Moore, W.S., 2000b. Determining coastal mixing rates using radium isotopes. Cont. Shelf Porcelli, D., Kim, C.-K., Martin, P., Moore, W.S., Phaneuf, M., 2014. The environmental Res. 20, 1993–2007. https://doi.org/10.1016/S0278-4343(00)00054-6. behaviour of radium: revised edition TRS No. 476. In: Behaviour of Radium : Revised Moore, W.S., 2003. Sources and fluxes of submarine groundwater discharge delineated Edition. IAEA, International Atomic Energy Agency, Vienna, pp. 6–32. by radium isotopes. Biogeochemistry 66, 75–93. https://doi.org/10.1023/B: Povinec, P.P., Burnett, W.C., Beck, A., Bokuniewicz, H., Charette, M., Gonneea, M.E., BIOG.0000006065.77764.a0. Groening, M., Ishitobi, T., Kontar, E., Kwong, L.L.W., Marie, D.E.P., Moore, W.S., Moore, W.S., 2006a. The role of submarine groundwater discharge in coastal Oberdorfer, J.A., Peterson, R., Ramessur, R., Rapaglia, J., Stieglitz, T., Top, Z., 2012. biogeochemistry. J. Geochem. Explor. 88, 389–393. https://doi.org/10.1016/j. Isotopic, geophysical and biogeochemical investigation of submarine groundwater gexplo.2005.08.082. discharge: IAEA-UNESCO intercomparison exercise at Mauritius Island. J. Environ. Moore, W.S., 2006b. Radium isotopes as tracers of submarine groundwater discharge in Radioact. 104, 24–45. https://doi.org/10.1016/j.jenvrad.2011.09.009. Sicily. Cont. Shelf Res. 26, 852–861. https://doi.org/10.1016/j.csr.2005.12.004. Prieto, C., Destouni, G., 2005. Quantifying hydrological and tidal influences on Moore, W.S., 2007. Seasonal distribution and flux of radium isotopes on the southeastern groundwater discharges into coastal waters. Water Resour. Res. 41, W12427. US continental shelf. J. Geophys. Res. 112, 1–16. https://doi.org/10.1029/ https://doi.org/10.1029/2004wr003920. 2007jc004199. Puigdomenech, I., Bergstrom, U., 1995. Calculation of distribution coefficients for Moore, W.S., 2008. Fifteen years experience in measuring 224Ra and 223Ra by delayed- radionuclides in soils and sediments. Nucl. Saf. 36, 142–154. coincidence counting. Mar. Chem. 109, 188–197. https://doi.org/10.1016/j. Rahman, S., Tamborski, J.J., Charette, M.A., Cochran, J.K., 2019. Dissolved silica in the marchem.2007.06.015. subterranean estuary and the impact of submarine groundwater discharge on the Moore, W.S., 2010. The Effect of Submarine Groundwater Discharge on the Ocean. In: global marine silica budget. Mar. Chem. 208, 29–42. https://doi.org/10.1016/j. Annual Review of Marine Science, pp. 59–88. https://doi.org/10.1146/annurev- marchem.2018.11.006. marine-120308-081019. Rajaomahefasoa, R.E., Voltaggio, M., Rakotomandrindra, P.F., Ratsimbazafy, J.B., Moore, W.S., 2015. Inappropriate attempts to use distributions of 228Ra and 226Ra in Spadoni, M., Rakoto, H.A., 2019. Radium isotopes for groundwater age and coastal waters to model mixing and advection rates. Cont. Shelf Res. 105, 95–100. sustainability in the highland of Antananarivo, Madagascar. J. African Earth Sci. https://doi.org/10.1016/j.csr.2015.05.014. 156, 94–107. https://doi.org/10.1016/j.jafrearsci.2019.05.011. Moore, W.S., Arnold, R., 1996. Measurement of 223 Ra and 224 Ra in coastal waters Rama, Moore, W.S., 1984. Mechanism of transport of U-Th series radioisotopes from using a delayed coincidence counter. J. Geophys. Res. 101, 1321–1329. solids into ground water. Geochim. Cosmochim. Acta 48, 395–399. https://doi.org/ Moore, W.S., Dymond, J., 1991. Fluxes of 226Ra and barium in the Pacific Ocean: The 10.1016/0016-7037(84)90261-8. importance of boundary processes. Earth Planet. Sci. Lett. 107, 55–68. https://doi. Rama, Moore, W.S., 1996. Using the radium quartet for evaluating groundwater input org/10.1016/0012-821X(91)90043-H. and water exchange in salt marshes. Geochim. Cosmochim. Acta 60, 4645–4652. Moore, W.S., Reid, D.F., 1973. Extraction of radium from natural waters using https://doi.org/10.1016/S0016-7037(96)00289-X. manganese-impregnated acrylic fibers. J. Geophys. Res. 78, 8880–8886. https://doi. Rapaglia, J., Ferrarin, C., Zaggia, L., Moore, W.S., Umgiesser, G., Garcia-Solsona, E., org/10.1029/JC078i036p08880. Garcia-Orellana, J., Masque, P., 2010. Investigation of residence time and Moore, W.S., Shaw, T.J., 2008. Fluxes and behavior of radium isotopes, barium, and groundwater flux in Venice Lagoon: comparing radium isotope and hydrodynamical uranium in seven Southeastern US rivers and estuaries. Mar. Chem. 108, 236–254. models. J. Environ. Radioact. 101, 571–581. https://doi.org/10.1016/j. https://doi.org/10.1016/j.marchem.2007.03.004. jenvrad.2009.08.010. Moore, W.S., Astwood, H., Lindstrom, C., 1995. Radium isotopes in coastal waters on the Rapaglia, J., Koukoulas, S., Zaggia, L., Lichter, M., Manfe, G., Vafeidis, A.T., Manfé, G., Amazon shelf. Geochim. Cosmochim. Acta 59, 4285–4298. https://doi.org/10.1016/ Vafeidis, A.T., 2012. Quantification of submarine groundwater discharge and 0016-7037(95)00242-R. optimal radium sampling distribution in the Lesina Lagoon, Italy. J. Mar. Syst. 91, Moore, W.S., Kjerfve, B., Todd, J.F., 1998. Identification of rain-freshened plumes in the 11–19. https://doi.org/10.1016/j.jmarsys.2011.09.003. coastal ocean using Ra isotopes and Si. J. Geophys. Res. Ocean. 103, 7709–7717. Regan, H.M., Colyvan, M., Burgman, M.A., 2002. A taxonomy and treatment of https://doi.org/10.1029/98jc00246. uncertainty for ecology and conservation biology. Ecol. Appl. 12, 618–628. https:// Moore, W.S., Blanton, J.O., Joye, S.B., 2006. Estimates of flushing times, submarine doi.org/10.1890/1051-0761(2002)012[0618:ATATOU]2.0.CO;2. groundwater discharge, and nutrient fluxes to Okatee Estuary, South Carolina. Relyea, J.F., 1982. Theoretical and experimental considerations for the use of the column J. Geophys. Res. 111, 1–14. C09006. https://doi.org/10.1029/2005jc003041. method for determining retardation factors. Radioact Waste Manag. Nucl Fuel Cycle Moore, W.S., Sarmiento, J.L., Key, R.M., 2008. Submarine groundwater discharge 3, 151–166. revealed by Ra-228 distribution in the upper Atlantic Ocean. Nat. Geosci. 1, Rengarajan, R., Sarma, V.V.S.S., 2015. Submarine groundwater discharge and nutrient 309–311. https://doi.org/10.1038/ngeo183. addition to the coastal zone of the Godavari estuary. Mar. Chem. 172, 57–69. Moore, W.S., Humphries, M.S., Benitez-Nelson, C.R., Pillay, L., Higgs, C., 2019. Transport https://doi.org/10.1016/j.marchem.2015.03.008. of Radium and Nutrients Through Eastern South African Beaches. J. Geophys. Res. Robinson, C.E., Xin, P., Santos, I.R., Charette, M.A., Li, L., Barry, D.A., 2018. Ocean. 124, 2010–2027. https://doi.org/10.1029/2018JC014772. Groundwater dynamics in subterranean estuaries of coastal unconfined aquifers: Nathwani, J.S., Phillips, C.R., 1979. Adsorption of 226Ra by soils (I). Chemosphere 8, Controls on submarine groundwater discharge and chemical inputs to the ocean. 285–291. https://doi.org/10.1016/0045-6535(79)90111-5. Adv. Water Resour. 115, 315–331. https://doi.org/10.1016/j. Neff, J.M., 2002. Bioaccumulation in Marine Organisms: Effect of Contaminants from Oil advwatres.2017.10.041. Well Produced Water. Elsevier. Rocha, C., Robinson, C.E., Santos, I.R., Waska, H., Michael, H.A., Bokuniewicz, H.J., Neff, J.M., Sauer, T.C., 1995. Barium in Produced Water: FATE and Effects in the Marine 2021. A place for subterranean estuaries in the coastal zone. Estuar. Coast. Shelf Sci. Environment. Washington, D.C. 250, 107167. https://doi.org/10.1016/j.ecss.2021.107167. Neuholz, R., Schnetger, B., Kleint, C., Koschinsky, A., Lettmann, K., Sander, S., Türke, A., Rodellas, R., Garcia-Orellana, J., Garcia-Solsona, E., Masqué, P., Domínguez, J.A., Walter, M., Zitoun, R., Brumsack, H.J., 2020. Near-field hydrothermal plume Ballesteros, B., Mejías, M., 2012. Quantifying groundwater discharge from different dynamics at Brothers Volcano (Kermadec Arc): A short-lived radium isotope study. sources into a Mediterranean wetland by using 222Rn and Ra isotopes. J. Hydrol. Chem. Geol. 533, 119379. https://doi.org/10.1016/j.chemgeo.2019.119379. 466–467. Oehler, T., Tamborski, J., Rahman, S., Moosdorf, N., Ahrens, J., Mori, C., Neuholz, R., Rodellas, V., Garcia-Orellana, J., Tovar-Sánchez, A., Basterretxea, G., López-Garcia, J.M., Schnetger, B., Beck, M., 2019. DSi as a Tracer for Submarine Groundwater Sánchez-Quilez, D., Garcia-Solsona, E., Masqué, P., 2014. Submarine groundwater Discharge. Front. Mar. Sci. 6, 1–13. https://doi.org/10.3389/fmars.2019.00563. discharge as a source of nutrients and trace metals in a Mediterranean bay (Palma Okubo, A., 1976. Remarks on the use of ‘diffusion diagrams’ in modeling scale- Beach, Balearic Islands). Mar. Chem. 160, 56–66. dependent diffusion. Deep-Sea Res. Oceanogr. Abstr. 23, 1213–1214. https://doi. Rodellas, V., Garcia-Orellana, J., Masqué, P., Feldman, M., Weinstein, Y., 2015a. org/10.1016/0011-7471(76)90897-4. Submarine groundwater discharge as a major source of nutrients to the Ollivier, P., Claude, C., Radakovitch, O., Hamelin, B., 2008. TIMS measurements of Ra- Mediterranean Sea. Proc. Natl. Acad. Sci. 112, 3926–3930. https://doi.org/10.1073/ 226 and Ra-228 in the Gulf of Lion, an attempt to quantify submarine groundwater pnas.1419049112. discharge. Mar. Chem. 109, 337–354. https://doi.org/10.1016/j. Rodellas, V., Garcia-Orellana, J., Masqué, P., Font-Muñoz, J.S., 2015b. The influence of marchem.2007.08.006. sediment sources on radium-derived estimates of Submarine Groundwater Paytan, A., Boehm, A.B., Shellenbarger, G.G., 2004. Bacterial Contamination and Discharge. Mar. Chem. 171, 107–117. https://doi.org/10.1016/j. Submarine Groundwater DischargeA Possible Link. Environ. Chem. 1, 29–30. marchem.2015.02.010. Paytan, A., Shellenbarger, G.G., Street, J.H., Gonneea, M.E., Davis, K., Young, M.B., Rodellas, V., Garcia-Orellana, J., Trezzi, G., Masqué, P., Stieglitz, T.C., Bokuniewicz, H., Moore, W.S., 2006. Submarine groundwater discharge: An important source of new Kirk Cochran, J., Berdalet, E., 2017. Using the radium quartet to quantify submarine inorganic nitrogen to coral reef ecosystems. Limnol. Oceanogr. 51, 343–348. groundwater discharge and porewater exchange. Geochim. Cosmochim. Acta 196, Petermann, E., Knöller, K., Rocha, C., Scholten, J., Stollberg, R., Weiß, H., Schubert, M., 58–73. https://doi.org/10.1016/j.gca.2016.09.016. 2018. Coupling End-Member Mixing Analysis and Isotope Mass Balancing (222-Rn) Rodellas, V., Stieglitz, T.C., Andrisoa, A., Cook, P.G., Raimbault, P., Tamborski, J.J., van for Differentiation of Fresh and Recirculated Submarine Groundwater Discharge Into Beek, P., Radakovitch, O., 2018. Groundwater-driven nutrient inputs to coastal Knysna Estuary, South Africa. J. Geophys. Res. Ocean. 123, 952–970. https://doi. lagoons: The relevance of lagoon water recirculation as a conveyor of dissolved org/10.1002/2017JC013008. nutrients. Sci. Total Environ. 642, 764–780. https://doi.org/10.1016/j. Peterson, R.N., Burnett, W.C., Dimova, N., Santos, I.R., 2009. Comparison of scitotenv.2018.06.095. measurement methods for radium-226 on manganese-fiber. Limnol. Oceanogr. Rodellas, V., Cook, P.G., Mccallum, J., Andrisoa, A., Stieglitz, T.C., 2020. Temporal Methods 7, 196–205. https://doi.org/10.4319/lom.2009.7.196. variations in porewater fluxes to a coastal lagoon driven by wind waves and changes Porcelli, D., Swarzenski, P.W., 2003. The behavior of U- and Th-series nuclides in in lagoon water depths. J. Hydrol. 124363 https://doi.org/10.1016/j. groundwater. In: Bourdon, B., Henderson, G.M., Lundstorm, C.C., Turner, S.P. (Eds.), jhydrol.2019.124363. Uranium- Series Geochemistry. The Mineralogical Society ofAmerica, Washington, Rodellas, V., Stieglitz, T.C., Tamborski, J.J., van Beek, P., Andrisoa, A., Cook, P.G., 2021. D.C., pp. 317–361 Conceptual uncertainties in groundwater and porewater fluxes estimated by radon 24 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 and radium mass balances. Limnol. Oceanogr. 1–19 https://doi.org/10.1002/ Swarzenski, P.W., 2007. U/Th series radionuclides as coastal groundwater tracers. Chem. lno.11678. Rev. 107, 663–674. https://doi.org/10.1021/cr0503761. Rosenberry, D.O., Duque, C., Lee, D.R., 2020. History and evolution of seepage meters for Swarzenski, P.W., Burnett, W.C., Greenwood, W.J., Herut, B., Peterson, R., Dimova, N., quantifying flow between groundwater and surface water: Part 1 – Freshwater Shalem, Y., Yechieli, Y., Weinstein, Y., 2006. Combined time-series resistivity and settings. Earth-Science Rev. 204, 103167. https://doi.org/10.1016/j. geochemical tracer techniques to examine submarine groundwater discharge at Dor earscirev.2020.103167. Beach, Israel. Geophys. Res. Lett 33, L24405. https://doi.org/10.1029/ Ruiz-González, C., Rodellas, V., Garcia-Orellana, J., 2021. The microbial dimension of 2006gl028282. submarine groundwater discharge: current challenges and future directions. FEMS Swarzenski, P.W., Reich, C., Kroeger, K.D., Baskaran, M., 2007a. Ra and Rn isotopes as Microbiol. Rev. 1–25. https://doi.org/10.1093/femsre/fuab010. natural tracers of submarine groundwater discharge in Tampa Bay, Florida. Mar. Rutgers van der Loeff, M., Kipp, L., Charette, M.A., Moore, W.S., Black, E., Stimac, I., Chem. 104, 69–84. https://doi.org/10.1016/j.marchem.2006.08.001. Charkin, A., Bauch, D., Valk, O., Karcher, M., Krumpen, T., Casacuberta, N., Swarzenski, P.W., Simonds, F.W., Paulson, A.J., Kruse, S., Reich, C., 2007b. Geochemical Smethie, W., Rember, R., 2018. Radium Isotopes across the arctic ocean show time and geophysical examination of submarine groundwater discharge and associated scales of water mass ventilation and increasing shelf inputs. J. Geophys. Res. Ocean. nutrient loading estimates into Lynch Cove, Hood Canal, WA. Environ. Sci. Technol. 123, 4853–4873. https://doi.org/10.1029/2018JC013888. 41, 7022–7029. https://doi.org/10.1021/es070881a. Sanial, V., van Beek, P., Lansard, B., Souhaut, M., Kestenare, E., D’Ovidio, F., Zhou, M., Szabo, B.J., 1967. Radium content in plankton and sea water in the Bahamas. Geochim. Blain, S., 2015. Use of Ra isotopes to deduce rapid transfer of sediment-derived Cosmochim. Acta 31, 1321–1331. https://doi.org/10.1016/S0016-7037(67)80018- inputs off Kerguelen. Biogeosci. Discuss. 12, 1415–1430. https://doi.org/10.5194/ 8. bg-12-1415-2015. Tachi, Y., Shibutani, T., Sato, H., Yui, M., 2001. Experimental and modeling studies on Santos, I.R., Lechuga-Deveze, C., Peterson, R.N., Burnett, W.C., 2011. Tracing submarine sorption and diffusion of radium in bentonite. J. Contam. Hydrol. 47, 171–186. hydrothermal inputs into a coastal bay in Baja California using radon. Chem. Geol. https://doi.org/10.1016/S0169-7722(00)00147-9. 282, 1–10. https://doi.org/10.1016/j.chemgeo.2010.12.024. Tagliabue, A., Aumont, O., Bopp, L., 2014. of Iron on the Global Carbon Cycle. Geophys. Santos, I.R., Chen, X., Lecher, A.L., Sawyer, A.H., Moosdorf, N., Rodellas, V., Res. Lett. 318, 920–926. https://doi.org/10.1002/2013GL059059.Received. Tamborski, J.J., Cho, H.-M., Dimova, N., Sugimoto, R., Bonaglia, S., Li, H., Tamborski, J.J., Rogers, A.D., Bokuniewicz, H.J., Cochran, J.K., Young, C.R., 2015. Hajati, M.-C., Li, L., 2021. Submarine groundwater discharge impacts on coastal Identification and quantification of diffuse fresh submarine groundwater discharge nutrient biogeochemistry. Nat. Rev. Earth Environ. 0123456789 https://doi.org/ via airborne thermal infrared remote sensing. Remote Sens. Environ. 171. https:// 10.1038/s43017-021-00152-0. doi.org/10.1016/j.rse.2015.10.010. Santos, I.R., Eyre, B.D., Huettel, M., 2012. The driving forces of porewater and Tamborski, J.J., Cochran, J.K., Bokuniewicz, H.J., 2017a. Submarine groundwater groundwater flow in permeable coastal sediments: A review. Estuar. Coast. Shelf Sci. discharge driven nitrogen fluxes to Long Island Sound, NY: Terrestrial vs. marine 98, 1–15. https://doi.org/10.1016/j.ecss.2011.10.024. sources. Geochim. Cosmochim. Acta 218, 40–57. https://doi.org/10.1016/j. Santos, I.R., Chen, X., Lecher, A.L., Sawyer, A.H., Moosdorf, N., 2021. Submarine gca.2017.09.003. groundwater discharge impacts on coastal nutrient biogeochemistry. Nat. Rev. Earth Tamborski, J.J., Cochran, J.K., Bokuniewicz, H.J., 2017b. Application of224Ra Environ. 0123456789 https://doi.org/10.1038/s43017-021-00152-0. and222Rn for evaluating seawater residence times in a tidal subterranean estuary. Sawyer, A.H., Shi, F., Kirby, J.T., Michael, H.A., 2013. Dynamic response of surface Mar. Chem. 189, 32–45. https://doi.org/10.1016/j.marchem.2016.12.006. water-groundwater exchange to currents, tides, and waves in a shallow estuary. Tamborski, J.J., Cochran, J.K., Heilbrun, C., Rafferty, P., Fitzgerald, P., Zhu, Q., J. Geophys. Res. Ocean. 118, 1749–1758. https://doi.org/10.1002/jgrc.20154. Salazar, C., 2017c. Investigation of pore water residence times and drainage Schmidt, C., Hanfland, C., Regnier, P., Van Cappellen, P., Schlueter, M., Knauthe, U., velocities in salt marshes using short-lived radium isotopes. Mar. Chem. 196, Stimac, I., Geibert, W., 2011. Ra-228, Ra-226, Ra-224 and Ra-223 in potential 107–115. https://doi.org/10.1016/j.marchem.2017.08.007. sources and sinks of land-derived material in the German Bight of the North Sea: Tamborski, J.J., Bejannin, S., Garcia-Orellana, J., Souhaut, M., Charbonnier, C., implications for the use of radium as a tracer. Geo-Marine Lett. 31, 259–269. https:// Anschutz, P., Pujo-Pay, M., Conan, P., Crispi, O., Monnin, C., Stieglitz, T., doi.org/10.1007/s00367-011-0231-5. Rodellas, V., Andrisoa, A., Claude, C., van Beek, P., 2018. A comparison between Schubert, M., Knoeller, K., Rocha, C., Einsiedl, F., 2015. Evaluation and source water recirculation and terrestrially-driven dissolved silica fluxes to the attribution of freshwater contributions to Kinvarra Bay, Ireland, using 222Rn, EC and Mediterranean Sea traced using radium isotopes. Geochim. Cosmochim. Acta 238, stable isotopes as natural indicators. Environ. Monit. Assess. 187, 1–15. https://doi. 496–515. https://doi.org/10.1016/j.gca.2018.07.022. org/10.1007/s10661-015-4274-3. Tamborski, J., van Beek, P., Rodellas, V., Monnin, C., Bergsma, E., Stieglitz, T., Selzer, S., Annett, A.L., Homoky, W.B., 2021. RaDeCC Reader: Fast, accurate and Heilbrun, C., Cochran, J.K., Charbonnier, C., Anschutz, P., Bejannin, S., Beck, A., automated data processing for Radium Delayed Coincidence Counting systems. 2019. Temporal variability of lagoon–sea water exchange and seawater circulation Comput. Geosci. 149 https://doi.org/10.1016/j.cageo.2021.104699. through a Mediterranean barrier beach. Limnol. Oceanogr. 64, 2059–2080. https:// Semkow, T.M., 1990. Recoil-emanation theory applied to radon release from mineral doi.org/10.1002/lno.11169. grains. Geochim. Cosmochim. Acta 54, 425–440. https://doi.org/10.1016/0016- Tamborski, J., Van Beek, P., Conan, P., Pujo-pay, M., Odobel, C., Ghiglione, J., Seidel, J., 7037(90)90331-E. Ar, B., Diego-Feliu, M., Garcia-Orellana, J., Szafran, A., Souhaut, M., 2020a. Sheldon, J.E., Alber, M., 2008. The calculation of estuarine turnover times using Submarine karstic springs as a source of nutrients and bioactive trace metals for the freshwater fraction and tidal prism models: a critical evaluation. Estu 29, 133–146. oligotrophic Northwest Mediterranean Sea. Sci. Total Environ. 732, 1–14. https:// Shellenbarger, G.G., Monismith, S.G., Genin, A., Paytan, A., 2006. The importance of doi.org/10.1016/j.scitotenv.2020.139106. submarine groundwater discharge to the nearshore nutrient supply in the Gulf of Tamborski, J.J., Cochran, J.K., Bokuniewicz, H., Heilbrun, C., Garcia-Orellana, J., Aqaba (Israel). Limnol. Oceanogr. 51, 1876–1886. https://doi.org/10.4319/ Rodellas, V., Wilson, R., 2020b. Radium mass balance sensitivity analysis for lo.2006.51.4.1876. submarine groundwater discharge estimation in semi-enclosed basins: The case Shi, X., Mason, R.P., Charette, M.A., Mazrui, N.M., Cai, P., 2018. Mercury flux from salt study of Long Island Sound. Front. Environ. Sci. 8, 108. https://doi.org/10.3389/ marsh sediments: Insights from a comparison between 224Ra/228Th disequilibrium FENVS.2020.00108. and core incubation methods. Geochim. Cosmochim. Acta 222, 569–583. https:// Tamborski, J.J., Eagle, M., Kurylyk, B.L., Kroeger, K.D., Wang, Z.A., Henderson, P., doi.org/10.1016/j.gca.2017.10.033. Charette, M.A., 2021. Pore water exchange-driven inorganic carbon export from Shi, X., Benitez-Nelson, C.R., Cai, P., He, L., Moore, W.S., 2019a. Development of a two- intertidal salt marshes. Limnol. Oceanogr. 1–19 https://doi.org/10.1002/lno.11721. layer transport model in layered muddy–permeable marsh sediments using Taniguchi, M., Burnett, W.C., Cable, J.E., Turner, J.V., 2002. Investigation of submarine 224Ra–228Th disequilibria. Limnol. Oceanogr. 64, 1672–1687. https://doi.org/ groundwater discharge. Hydrol. Process. 16, 2115–2129. https://doi.org/10.1002/ 10.1002/lno.11143. hyp.1145. Shi, Xiangming, Wei, L., Hong, Q., Liu, L., Wang, Y., Shi, Xueying, Ye, Y., Cai, P., 2019b. Taniguchi, M., Dulai, H., Burnett, K.M., Santos, I.R., Sugimoto, R., Stieglitz, T., Kim, G., Large benthic fluxes of dissolved iron in China coastal seas revealed by 224Ra/ Moosdorf, N., Burnett, W.C., 2019. Submarine Groundwater Discharge: Updates on 228Th disequilibria. Geochim. Cosmochim. Acta 260, 49–61. https://doi.org/ Its Measurement Techniques, Geophysical Drivers, Magnitudes, and Effects. Front. 10.1016/j.gca.2019.06.026. Environ. Sci. 7, 1–26. https://doi.org/10.3389/fenvs.2019.00141. Simpson, G.C., Wright, C.S., Schuster, A., 1911. Atmospheric electricity over the ocean. Te Hsieh, Y., Henderson, G.M., 2011. Precise measurement of 228Ra/226Ra ratios and Proc. R. Soc. London. Ser. A, Contain. Pap. a Math. Phys. Character 85, 175–199. Ra concentrations in seawater samples by multi-collector ICP mass spectrometry. https://doi.org/10.1098/rspa.1911.0031. J. Anal. At. Spectrom. 26, 1338–1346. https://doi.org/10.1039/c1ja10013k. Snavely, E.S., 1989. Radionuclides in Produced Water. Washington, D.C. Tomasky-Holmes, G., Valiela, I., Charette, M.A., 2013. Determination of water mass ages Stieglitz, T.C., Clark, J.F., Hancock, G.J., 2013. The mangrove pump: The tidal flushing using radium isotopes as tracers: Implications for phytoplankton dynamics in of animal burrows in a tropical mangrove forest determined from radionuclide estuaries. Mar. Chem. 156, 18–26. https://doi.org/10.1016/j. budgets. Geochim. Cosmochim. Acta 102, 12–22. https://doi.org/10.1016/j. marchem.2013.02.002. gca.2012.10.033. Toth, J., 1963. A Theoretical Analysis of Groundwater Flow in Small Drainage Basins 1 of Sugimoto, R., Kitagawa, K., Nishi, S., Honda, H., Yamada, M., Kobayashi, S., Shoji, J., phe - low order stream and having similar t he outlet of lowest impounded body of a Ohsawa, S., Taniguchi, M., Tominaga, O., 2017. Phytoplankton primary productivity relatively. J. Geophys. Res. 68, 4795–4812. https://doi.org/10.1029/ around submarine groundwater discharge in nearshore coasts. Mar. Ecol. Prog. Ser. JZ068i016p04795. 563 https://doi.org/10.3354/meps11980. Tovar-Sánchez, A., Basterretxea, G., Rodellas, V., Sánchez-Quiles, D., García-Orellana, J., Sun, H., Semkow, T.M., 1998. Mobilization of thorium, radium and radon radionuclides Masqué, P., Jordi, A., López, J.M., Garcia-Solsona, E., 2014. Contribution of in ground water by successive alpha-recoils. J. Hydrol. 205, 126–136. https://doi. groundwater discharge to the coastal dissolved nutrients and trace metal org/10.1016/S0022-1694(97)00154-6. concentrations in Majorca Island: karstic vs detrital systems. Environ. Sci. Technol. Sun, Y., Torgersen, T., 2001. Adsorption-desorption reactions and bioturbation transport 48, 11819–11827. https://doi.org/10.1021/es502958t. of 224Ra in marine sediments: A one-dimensional model with applications. Mar. Trezzi, G., Garcia-Orellana, J., Santos-Echeandia, J., Rodellas, V., Garcia-Solsona, E., Chem. 74, 227–243. https://doi.org/10.1016/S0304-4203(01)00017-2. Garcia-Fernandez, G., Masqué, P., 2016. The influence of a metal-enriched mining 25 J. Garcia-Orellana et al. Earth-Science Reviews 220 (2021) 103681 waste deposit on submarine groundwater discharge to the coastal sea. Mar. Chem. counting system (RaDeCC). J. Environ. Radioact. 99, 1859–1862. https://doi.org/ 178, 35–45. https://doi.org/10.1016/j.marchem.2015.12.004. 10.1016/j.jenvrad.2008.08.008. Tricca, A., Wasserburg, G.J., Porcelli, D., Baskaran, M., 2001. The transport of U- and Th- Webster, I.T., Hancock, G.J., Murray, A.S., 1995. Modelling the effect of salinity on series nuclides in sandy confined aquifers. Geochim. Cosmochim. Acta 65, radium desorption from sediments. Geochim. Cosmochim. Acta 59, 2469–2476. 1187–1210. https://doi.org/10.1016/S0016-7037(02)01341-8. https://doi.org/10.1016/0016-7037(95)00141-7. Trier, R.M., Broecker, W.S., Feely, H.W., 1972. Radium-228 profile at the second geosecs Weinstein, Y., Burnett, W.C., Swarzenski, P.W., Shalem, Y., Yechieli, Y., Herut, B., 2007. intercalibration station, 1970, in the North Atlantic. Earth Planet. Sci. Lett. 16, Role of aquifer heterogeneity in fresh groundwater discharge and seawater 141–145. recycling: An example from the Carmel coast, Israe. J. Geophys. Res. 112, C12016. van Beek, P., François, R., Conte, M., Reyss, J.-L., Souhaut, M., Charette, M., 2007. https://doi.org/10.1029/2007jc004112. 228 Ra/226Ra and 226Ra/Ba ratios to track barite formation and transport in the water Weinstein, Y., Yechieli, Y., Shalem, Y., Burnett, W.C., Swarzenski, P.W., Herut, B., 2011. column. Geochim. Cosmochim. Acta 71, 71–86. https://doi.org/10.1016/j. What Is the Role of Fresh Groundwater and Recirculated Seawater in Conveying gca.2006.07.041. Nutrients to the Coastal Ocean? Environ. Sci. Technol. 45, 5195–5200. https://doi. Veeh, H.H., Moore, W.S., Smith, S.V., 1995. The behaviour of uranium and radium in an org/10.1021/es104394r. inverse estuary. Cont. Shelf Res. 15, 1569–1583. https://doi.org/10.1016/0278- Werner, A.D., Bakker, M., Post, V.E.A., Vandenbohede, A., Lu, C., Ataie-Ashtiani, B., 4343(95)00031-U. Simmons, C.T., Barry, D.A., 2013. Seawater intrusion processes, investigation and Vengosh, A., Hirschfeld, D., Vinson, D., Dwyer, G., Raanan, H., Rimawi, O., Al-Zoubi, A., management: Recent advances and future challenges. Adv. Water Resour. 51, 3–26. Akkawi, E., Marie, A., Haquin, G., Zaarur, S., Ganor, J., 2009. High naturally https://doi.org/10.1016/j.advwatres.2012.03.004. occurring radioactivity in fossil groundwater from the middle east. Environ. Sci. Willett, I.R., Bond, W.J., 1995. Sorption of manganese, uranium, and radium by highly Technol. 43, 1769–1775. https://doi.org/10.1021/es802969r. weathered soils. J. Environ. Qual. 24, 834–845. https://doi.org/10.2134/ Vieira, L.H., Achterberg, E.P., Scholten, J., Beck, A.J., Liebetrau, V., Mills, M.M., jeq1995.00472425002400050006x. Arrigo, K.R., 2019. Benthic fluxes of trace metals in the Chukchi Sea and their Young, M.B., Gonneea, M.E., Fong, D.A., Moore, W.S., Herrera-Silveira, J., Paytan, A., transport into the Arctic Ocean. Mar. Chem. 208, 43–55. https://doi.org/10.1016/j. 2008. Characterizing sources of groundwater to a tropical coastal lagoon in a karstic marchem.2018.11.001. area using radium isotopes and water chemistry. Mar. Chem. 109, 377–394. https:// Vieira, L.H., Krisch, S., Hopwood, M.J., Beck, A.J., Scholten, J., Liebetrau, V., doi.org/10.1016/j.marchem.2007.07.010. Achterberg, E.P., 2020. Unprecedented Fe delivery from the Congo River margin to Younger, P.L., 1996. Submarine groundwater discharge. Nature 382, 121–122. https:// the South Atlantic Gyre. Nat. Commun. 11, 1–8. https://doi.org/10.1038/s41467- doi.org/10.1038/382121a0. 019-14255-2. Zarroca, M., Linares, R., Rodellas, V., Garcia-Orellana, J., Roqué, C., Bach, J., Masqué, P., Vieira, L.H., Geibert, W., Stimac, I., Koehler, D., Rutgers van der Loeff, M.M., 2021. The 2014. Delineating coastal groundwater discharge processes in a wetland area by analysis of 226Ra in 1-liter seawater by isotope dilution via single-collector sector- means of electrical resistivity imaging, 224Ra and 222Rn. https://doi.org/10.1002/ field ICP-MS. Limnol. Oceanogr. Methods. https://doi.org/10.1002/lom3.10428. hyp.9793. Wang, X., Du, J., 2016. Submarine groundwater discharge into typical tropicallagoons: A Zhang, Y., Schaap, M.G., 2019. Estimation of saturated hydraulic conductivity with case study in eastern Hainan Island, China. Geochem. Geophys. Geosyst. 17, pedotransfer functions: A review. J. Hydrol. 575, 1011–1030. https://doi.org/ 4366–4382. https://doi.org/10.1002/2016GC006502. 10.1016/j.jhydrol.2019.05.058. Wang, X., Du, J., Ji, T., Wen, T., Liu, S., Zhang, J., 2014. An estimation of nutrient fluxes Zhang, Y., Li, H., Xiao, K., Wang, Xuejing, Lu, X., Zhang, M., An, A., Qu, W., Wan, L., via submarine groundwater discharge into the Sanggou Bay-A typical multi-species Zheng, C., Wang, Xusheng, Jiang, X., 2017. Improving Estimation of Submarine culture ecosystem in China. Mar. Chem. 167 https://doi.org/10.1016/j. Groundwater Discharge Using Radium and Radon Tracers: Application in Jiaozhou marchem.2014.07.002. Bay, China. J. Geophys. Res. Ocean. 122, 8263–8277. https://doi.org/10.1002/ Warner, J.C., Rockwell Geyer, W., Arango, H.G., 2010. Using a composite grid approach 2017JC013237. in a complex coastal domain to estimate estuarine residence time. Comput. Geosci. Zhou, Y.Q., Befus, K.M., Sawyer, A.H., David, C.H., 2018. Opportunities and Challenges 36, 921–935. https://doi.org/10.1016/j.cageo.2009.11.008. in Computing Fresh Groundwater Discharge to Continental Coastlines: A Multimodel Waska, H., Kim, S., Kim, G., Peterson, R.N., Burnett, W.C., 2008. An efficient and simple Comparison for the United States Gulf and Atlantic Coasts. Water Resour. Res. 54, method for measuring Ra-226 using the scintillation cell in a delayed coincidence 8363–8380. https://doi.org/10.1029/2018WR023126. 26