Modeling the Minho River plume intrusion into the Rias Baixas (NW Iberian Peninsula)

Continental Shelf Research 85 (2014) 30–41

Contents lists available at ScienceDirect

Continental Shelf Research journal homepage: www.elsevier.com/locate/csr

Modeling the Minho River plume intrusion into the Rias Baixas (NW Iberian Peninsula) M.C. Sousa a, N. Vaz a, I. Alvarez a,b, M. Gomez-Gesteira b, J.M. Dias a,n a b

CESAM, Departamento de Física, Universidade de Aveiro, 3810-193 Aveiro, Portugal EPhysLab (Environmental Physics Laboratory), Universidade de Vigo, Facultade de Ciencias, Ourense, Spain

art ic l e i nf o

a b s t r a c t

Article history: Received 15 May 2013 Received in revised form 8 May 2014 Accepted 10 June 2014 Available online 18 June 2014

The Minho River discharge is recognized as particularly important in driving the circulation and hydrography of Rias Baixas, which are highly productive fishery and aquaculture regions extremely sensitive to environmental characteristics. The intrusion of the Minho River plume inside these Rias can reverse the normal circulation pattern and affect the macronutrient concentrations, imposing a control on new production within the estuarine environment. Consequently, detailed knowledge of the propagation of the plume in this zone facilitates largely the management of many exploited and protected local species. Thus, the main purpose of this work is to study the propagation and influence of Minho estuarine plume in Rias Baixas circulation and hydrography through the development and validation of an application of MOHID numerical model including a local coastal nesting configuration fed by Minho River discharge predicted by an estuary model. The nesting configuration and the Minho estuary model were validated and then applied to research the role of the wind and Minho River discharge effects on the circulation reversal. The spring of 1998 was chosen as the validation period for Minho estuarine plume propagation, considering there were field data available for this period confirming the intrusion of the Minho River plume in Rias Baixas and reversing the normal circulation pattern. Predictions replicate accurately the hydrodynamics and thermohaline patterns in Minho estuary and Rias Baixas under these conditions showing that the developed model application reproduces the dynamics of the coupled estuarine-near coastal systems under research. Results showed that a buoyancy intrusion caused by the Minho River reverses the normal estuarine salinity longitudinal gradient and estuarine circulation of the Rias de Vigo and Pontevedra. Moreover, it was found that a continuous moderate Minho River discharge combined with southerly winds is enough to reverse the Rias Baixas circulation pattern, reducing the importance of the occurrence of specific events of high runoff values. & 2014 Elsevier Ltd. All rights reserved.

Keywords: Nested models River discharge Negative circulation MOHID Western Galician coast

1. Introduction Freshwater inputs from rivers have a great influence on coastal waters, changing the distribution of particulate and dissolved matter, pollutants, nutrients, biogeochemical and phytoplankton communities (Bruland et al., 2008; Dortch and Whitledge, 1992; Kortzinger, 2003; Reifel et al., 2009; Warrick and Milliman, 2003). Generally, river plumes are turbid and rich in nutrients, remaining near surface due to their buoyancy and breaking up into lenses of less saline water, stimulating phytoplankton growth (Lunven et al., 2005). Some studies suggest that areas corresponding to river plumes are preferred feeding places for zooplankton (Pearcy, 1992), showing the relevance of a detailed analysis of river discharges effects on coastal estuaries located north or south of

n

Corresponding author. E-mail address: [email protected] (J.M. Dias).

http://dx.doi.org/10.1016/j.csr.2014.06.004 0278-4343/& 2014 Elsevier Ltd. All rights reserved.

the river mouth. External sources of freshwater can affect the density inside the estuaries modifying the mean baroclinic circulation and the macronutrients concentrations available for production (Banas et al., 2004; Das et al., 2012; Wong and Lu, 1994). The inflow of nutrient-rich water can impose a control on new production within the estuarine environment (Monteiro and Largier, 1999). Several studies conducted over different coastal systems around the world revealed the major interest in studying the ecological consequences of freshwater intrusion on estuaries nearby (e.g. Columbia River Banas et al., 2004, Delaware River Wong and Lu, 1994 or Mississippi River Das et al., 2012). These studies have shown that under downwelling winds the plume from the Columbia River can produce currents in the surface layers along the Washington coast as large as the wind driven-currents, affecting the out-migrating juvenile salmon (Hickey et al., 1998; Pearcy, 1992). Moreover, high nutrient concentrations from Mississippi River led to eutrophication in stratified coastal waters of the northern Gulf of Mexico, which affects marsh plants causing

M.C. Sousa et al. / Continental Shelf Research 85 (2014) 30–41

faster decomposition of the soil (Das et al., 2012; Swarzenski et al., 2008). The plume dynamics is controlled by the interaction of outflow inertia (Horner-Devine et al., 2009), buoyancy forcing (Jurisa and Chant, 2012), Earth's rotation effect (Monteiro et al., 2011), wind forcing (Choi and Wilkin, 2007) and alongshore currents (Fong and Geyer, 2002). The relation of these factors results in a variety of behaviors and times scales of river plumes. Thus, different plume classifications have been carried out (Garvine, 1995; Lentz and Largier, 2006; Yankovsky and Chapman, 1997). For example, they established relationships between the plume extension, velocity of the discharged freshwater and the depth of the river mouth. Considering the importance and different temporal and spatial scales of all the driving factors the development and exploitation of coastal circulation models has become an important tool to study the evolution of coastal plumes. From their application, individual effects of these driving factors (e.g. river discharge, tide or wind direction) on a coastal plume may be evaluated under different conditions. Several numerical studies related to coastal plumes characterization have been performed worldwide (Choi and Wilkin, 2007; Fong and Geyer, 2002; García-Berdeal et al., 2002; Guo and Valle-Levinson, 2007; Horner-Devine et al., 2009; Jurisa and Chant, 2012; Otero et al., 2008), indicating the utility of numerical models to analyze their dynamics as well as the thermohaline patterns of the areas affected by its spreading. These studies showed that downwelling favorable wind compresses the plume toward the coast and vertical mixing reduces stratification in the plume, where velocities are high. On the other hand, during upwelling favorable winds, the main feature is offshore extension of the plume, increasing vertical stratification in the area under its influence. Moreover, tidal effects contribute to increase mixing processes, reducing salinity stratification. The western coast of NW Iberian Peninsula is characterized by the presence of four estuaries locally named “Rias Baixas”. This area is characterized by a high primary production mainly due to spring-summer upwelling events that can support high fishery and aquaculture yields in this region (Tenore et al., 1995). Indeed, the Galician area produces around 250,000 t of mussels per year, i.e. around 15% of world's production. Consequently, knowledge of freshwater effects in these areas greatly facilitates the management of many exploited and protected species. The Minho River, situated 30 km south of the Rias Baixas, is the most important freshwater source flowing into this coastal region. Its buoyancy can flood the Rias Baixas for long periods, reversing the normal estuarine density gradients (Alvarez et al., 2006; Fiedler and Laurs, 1990), and motivating the research of its influence on these estuaries (Alvarez et al., 2006; Mourino and Fraga, 1982; Sousa et al., 2011). Previous studies revealed that plume intrusion can generate an important salinity decrease at estuaries mouth (Mourino and Fraga, 1982; Otero et al., 2013; Sousa et al., 2011), reversing the normal circulation pattern, with near bed water moving seawards and near surface water moving landwards. This process tends to stop water exchange between estuaries and shelf (Alvarez et al., 2006; deCastro et al., 2006). In addition, biogeochemical patterns can be modified suggesting the existence of blooms penetrating the Ria from shelf, embedded in a water mass that is fresher than estuarine one. According to deCastro et al. (2006) the Minho River freshwater intrusion effects in Ria de Pontevedra are characterized by a peculiar pattern of nutrient salt and oxygen distribution, affecting the photic layer and the near surface oxygen saturation. This pattern results from the surface water mass originated from mixing between coastal water and low-salinity nutrient-rich freshwater from the Minho River. The present literature review showed that the Minho River plume influence inside the western Galician estuaries has not been

31

previously studied in detail, namely by means of numerical model applications specifically designed for this purpose. Thus, this paper aims to study the propagation and influence of Minho estuarine plume in Rias Baixas circulation and hydrography. Additionally, it describes the implementation and validation of a nesting numerical modeling methodology developed to reproduce the propagation of the Minho estuarine plume toward the Rias Baixas, using a downscaling approach and a local estuarine model. In this context, the models ability to reproduce the Minho River plume intrusion into the Rias Baixas is analyzed for the spring of 1998, considering the availability of in situ data for this period confirming the presence of the Minho River plume inside these estuaries inducing estuarine circulation reversal. A high Minho River discharge and favorable wind patterns to advect the river plume toward the Rias Baixas were reported during this period (Alvarez et al., 2006). Finally, the individual possible effect of this high discharge and wind in Rias Baixas circulation and hydrography is also analyzed simulating several forcing scenarios.

2. Study area 2.1. Minho estuary The Minho River is about 300 km long, having a south-east direction alignment in the boundary between Portugal and Spain (Iberian Peninsula) (Fig. 1b). The river has a catchment area of 17,080 km2 and an annual average discharge of 300 m3 s
1. Its monthly average discharge oscillates between 100 m3 s
1 in August and 800 m3 s
1 in February. The Minho estuary (Fig. 1c) is approximately 38 km long with a total area of 23 km2. The estuary has a maximum width of about 2 km near the mouth, decreasing to about 10 m at the head. Estuary mean depth is 2.6 m and the maximum depth is about 4 m near the mouth (Freitas et al., 2009). It presents a semidiurnal, high-mesotidal regime and the range of the astronomical tide varies between 2 m, during neap tides, and almost 4 m, in spring tides (IH, 2006). The estuary can be considered as a partially mixed system, although it may present salt wedge characteristics during high river flow events (Sousa et al., 2005). The limit of salt intrusion is about 35 km from the mouth (Bettencourt et al., 2003).

2.2. Rias Baixas The Rias Baixas are four estuaries located south of Cape Finisterra, along the northwest Atlantic coast of the Iberian Peninsula (Fig. 1b). The Rias Baixas considered in this study are, from south to north, Ria de Vigo, Ria de Pontevedra and Ria de Arousa (Fig. 1b). They are connected to the open sea by means of two entrances defined by sets of islands in the outermost area. Freshwater contributions come from four small rivers: Oitaben–Verdugo River at the Ria de Vigo head, Lérez River at the Ria de Pontevedra and Umia and Ulla Rivers at the Ria de Arousa. In the Rias Baixas tidal forcing is mainly semidiurnal, with a significant low form number (lower than 0.25) (Varela et al., 2005). The Rias are mesotidal, with a tidal range from 2 m to 4 m (Fraga and Margalef, 1979), and generally behave as partially mixed estuaries with positive residual circulation, showing a two layer pattern with surface water outflow and bottom water inflow (Fraga and Margalef, 1979; Prego and Fraga, 1992). Under southerly winds and high Minho River discharges (1300 m3 s
1), this behavior is reversed, with water near bed moving seawards and water near surface moving landwards (deCastro et al., 2004, 2006).

32

M.C. Sousa et al. / Continental Shelf Research 85 (2014) 30–41

Fig. 1. Nested grids (a), Rias Baixas (b) and Minho estuary (c) bathymetries with indication of sampling stations position (black dots), tide gauges (black diamonds) and the location of the section (black rectangle) used to compute the Minho River outflow. The n, s and i indices correspond to the northern mouth, southern mouth and innermiddle areas of the Rias, respectively.

3. Model

3.1. Model physics

To achieve the main objectives of this research a model application composed by a three level nested model (Fig. 1a) is designed. This application is used to simulate the propagation of the Minho River plume toward the Rias Baixas and analyze the effects of the plume intrusion into these estuaries. A model application for the Minho estuary was also developed (Fig. 1c). It runs offline and was developed to reproduce the Minho estuaryocean interaction, computing the estuarine outflow which is introduced as a point discharge in the third level of the nested application.

The model application developed in this study is based in MOHID (www.mohid.com), a three-dimensional free surface numerical model which solves the Reynolds averaged form of the Navier–Stokes equations, which are discretized using a finite volume approach in a structured grid. The discrete form of the governing equations is applied macroscopically to a cell control volume. This method makes the solution independent of mesh geometry, allowing the use of a generic vertical mesh. Equations are discretized horizontally using an Arakawa C staggered grid and temporal discretization is performed by a semi-implicit (ADI)

M.C. Sousa et al. / Continental Shelf Research 85 (2014) 30–41

algorithm with two time levels per iteration. In this work, the horizontal and vertical advection of momentum and mass are computed using a TVD-Superbee method. Vertical turbulent mixing is computed using k–ε model (Canuto et al., 2001) included in the General Ocean Turbulence Model (GOTM) (Burchard et al., 1998). This model has the ability to simulate flows in shallow systems, like the Minho estuary, as well as to study the Western Iberian coastal circulation, and it has been previously applied to simulate the Galician Rias (Gomez-Gesteira et al., 1999; RuizVillarreal et al., 2002; Taboada et al., 1998). A more detailed description of the numerical algorithms can be found in Leitão (2002).

3.2. Configuration of the coastal nested model including the Rias Baixas A downscaling methodology to study the dynamics of the Minho estuarine plume was developed, similar to that proposed by Leitão et al. (2005) and Vaz et al. (2009b) for simulating the Algarve coastal circulation and Tagus estuarine plume. Here a three level one-way nested model configuration was implemented (Fig. 1a and b). This nested model comprises a large domain, used to compute the barotropic tide (L1) and two smaller baroclinic domains (L2 and L3), which are used to simulate estuarine plume advection. The main difference between L2 and L3 is horizontal resolution, which is coarser for L2. The downscaling approach was implemented to smooth the solution transition between L1 and L3. The first domain (L1, Fig. 1a) ranges from 13.51W to 11E and 33.51N to 501N, with a horizontal resolution of 0.061 and it was constructed based on the ETOPO1 global database. This domain is a 2D barotropic tidal driven model only using the FES2004 (Finite Element Solution) global solution as forcing (Lyard et al., 2006). The time step is 180 s and the horizontal eddy viscosity is 100 m2 s
1. For the levels, at the open boundary a radiation boundary scheme was used (Blumberg and Kantha, 1985). The ocean boundary conditions are given in cascade starting at the first level. Therefore, the 2D barotropic model is only used to predict the external tidal conditions necessary to feed the L2 baroclinic model with surface elevation. The second domain (L2, Fig. 1a) comprises a region from 10.081W to 8.401W and 40.921N to 43.501N with a horizontal resolution of 0.021. The third domain (L3, Fig. 1b) is from 9.521W to 8.601W and 41.681N to 42.861N with a horizontal step of 0.0051 and includes the Rias Baixas adjacent coastal area and it is directly coupled to L2 at the open boundaries. Higher resolution is used in order to properly simulate the small scale processes inside the Rias Baixas and also the Minho estuarine plume dispersion. L2 and L3 bathymetries were constructed based on the General Bathymetric Chart of the Oceans (GEBCO), with some corrections on continental shelf. A z-level vertical discretization was adopted, with L2 and L3 having a maximum of 46 and 42 vertical layers, respectively. To obtain the initial ocean stratification, L2 and L3 are forced at the open boundaries with the salinity and water temperature monthly mean profiles climatologies from Levitus (http://www. nodc.noaa.gov/OC5/WOA09/pr_woa09.html) (Antonov et al., 2010; Locarnini et al., 2010). In order to ensure model stability, second (third) level uses a time step of 60 s (15 s) and a turbulent horizontal eddy viscosity inside the domain of 20 m2 s
1 (5 m2 s
1). Furthermore, in both levels (L2 and L3), the zonal and meridional velocity components, salinity and water temperature in boundary cells are relaxed from the previous level down. For relaxation, the Flow Relaxation Scheme (Martinsen and Engedahl, 1987), which consists in applying a relaxation scheme at the boundary with an extension of ten cells is activated. The baroclinic force is slowly active over 10 inertia periods. The biharmonic filter

33

coefficients are set to 1  107 m4 s
1 and 1  104 m4 s
1 for L2 and L3, respectively. The surface boundary condition is imposed using the high resolution results from the Weather Research and Forecasting Model (WRF) with a spatial resolution of 4 km. These fields were interpolated into hourly fields for the two last model domains using triangulation interpolation in space and linear interpolation in time. At the surface, the sensible and latent heat fluxes are calculated using the Bowen and Dalton laws, respectively (Chapra, 1997). For bottom boundary condition, shear friction stress was imposed, assuming a velocity logarithmic profile. 3D momentum (zonal and meridional velocities), heat and salt balance equations are computed implicitly in the vertical direction, while in the horizontal directions are computed explicitly. As landward boundary condition (L3), freshwater input from Rias Baixas and the Minho estuarine outflow are considered. The Oitaben–Verdugo, Lérez, Umia and Ulla River discharges were obtained from estimations presented by Otero et al. (2010). The Minho estuarine outflow is imposed offline in L3. 3.3. The Minho estuary model A precise reproduction of the effect of the tidally dominated Minho estuary on the adjacent coast should take into account the tidal modulation of estuarine outflow fluxes and the accurate specification of salinity and water temperature at the estuary mouth. Consequently the Minho estuarine outflow must be determined using a local estuary model developed for this purpose. The use of a 1D model would be a possibility, but this approach would not take into account the spatial variability of the tidal propagation inside the estuary. Considering that the use of 2D depth integrated models is generalized in the scientific community to explain circulation features of shallow estuarine systems as Minho (e.g.: Plus et al., 2009; Santoro et al., 2011; Ouellet et al., 2013), in this study was decided to use this type of approach. This estuarine application runs separately to predict momentum, water and mass discharge, which are imposed to the three level nested models (L3 model). Thus, in L3 model takes into account the runoff, as well as the salinity and water temperature varying with the tide. This model was implemented in a 2D mode, due to the low depth of the estuary, and also taking into account the specific objective of this implementation (to reproduce the most realistic possible the estuarine water properties at the mouth of the estuary under different discharge conditions), as well as to maintain computational efficiency. In this configuration, a variable spatial step grid was developed due to the geometry of the estuary, which is characterized by several sand banks and islands. The grid has 119  100 cells, with dimensions of 100 m in the inner part of the estuary and 650 m (300 m in the direction y) at the western boundary. Numerical bathymetry was interpolated from topo-hydrographic data measured by the Hydrographic Institute of the Portuguese Navy in 1978/1989/1999 (Fig. 1c). The model is forced by tides at offshore open boundary and by river flow at upstream end of the Minho River estuary. Tidal forcing at oceanic open boundary is specified using a global tidal model (Le Provost et al., 1998). The Minho River freshwater input was supplied by the “Confederación Hidrográfica del Miño-Sil”. At ocean and river boundaries, the water temperature and salinity are considered fixed, with typical values for the season of simulation. At surface, heat fluxes were imposed, using latent and sensible heat fluxes parameterizations based on the Dalton and Bowen laws, respectively (Chapra, 1997). Meteorological data used for heat fluxes calculation were obtained from MeteoGalicia (www.meteogalicia.es). The time step defined for this application is 10 s, horizontal eddy viscosity is 10 m2 s
1, and a constant value of 0.0025 is assumed for bottom rugosity. Initial conditions for the hydrodynamic model are null free

34

M.C. Sousa et al. / Continental Shelf Research 85 (2014) 30–41

surface gradient and null velocity in all grid points. The model is used in this study to compute the estuary outflow along the section shown in Fig. 1c (black rectangle). Hourly discharges flows were determined using the current magnitude, depth and water level on the cells that define the section. Salinity and water temperature along the section were determined averaging their values for these cells.

4. Results and discussion 4.1. Model validation Two sets of simulations were performed to validate estuarine and coastal models (L3) settings in reproducing estuarine dynamics and fate of the estuarine plume originated from the Minho freshwater discharge. 4.1.1. Minho estuary Sea surface elevation (SSE), salinity and water temperature data sampled at the Minho estuary during March 2006 by the Hydrographic Institute of Portugal under the project “Estuarine Contributions to Inner Shelf Dynamics (ECOIS)” were used to validate the 2D Minho estuarine model. The set of simulations of the Minho estuarine model covers the period between February (model spin-up) and March 2006, but only results from March 2006 were used for validation. In order to quantify model accuracy in reproducing in situ tidal data, root mean square (RMS) and predictive skill (Warner et al., 2005) at Barra (41152.00 N, 8151.20 W), Caminha (41152.40 N, 8150.50 W) and Seixas (41154.00 N, 8148.20 W) (Fig. 1c) are computed following the methodology proposed by Sousa and Dias (2007) and Dias et al. (2009). The average RMS and the predictive skill are 0.16 m and 0.96 m, respectively. These values are similar to those obtained in previous numerical modeling works for other estuaries, such as the Ria de Aveiro (Sousa and Dias, 2007; Vaz et al., 2009a) and the Ria Formosa (Dias et al., 2009), showing that the estuarine model accuracy is excellent. Comparison between harmonic constants for major constituents computed from model predictions and observations is another quantification method used to perform the evaluation of the model accuracy. This methodology was applied in this study, by comparing harmonic constants of M2, S2, K1 and O1 constituents in the Minho estuary for stations shown in Fig. 1c (Table 1). Results show that tide is semi-diurnal with low diurnal inequality, with form number of about 0.08. The agreement between predicted and observed values is good both in amplitude and in phase for both semidiurnal constituents. The difference between predicted and observed amplitudes is between 0.06 m Table 1 Harmonic analysis results comparison of observed and predicted sea surface elevation data for Barra, Caminha and Seixas (M2, S2, O1 and K1 constituents). Tide gauge

M2

S2

O1

K1

Barra Caminha Seixas Barra Caminha Seixas Barra Caminha Seixas Barra Caminha Seixas

Amplitude (m)

Phase (1)

Data

Model

Difference

Data

Model

Difference

0.98 0.87 0.72 0.43 0.37 0.26 0.05 0.06 0.04 0.05 0.05 0.07

1.04 0.98 0.84 0.49 0.45 0.36 0.06 0.06 0.05 0.05 0.05 0.05


0.06
0.11
0.12
0.06
0.08
0.10
0.01 0.00
0.01 0.00 0.00 0.03

84.74 95.62 106.87 122.55 137.91 156.07 324.53 338.96 325.60 64.17 94.28 93.57

80.65 80.94 99.63 114.97 116.30 139.98 320.03 324.47 341.11 79.17 84.11 100.42

4.09 14.68 7.24 7.58 21.61 16.09 4.50 14.49
15.51
15.00 10.17
6.85

and 0.12 m for Barra and Seixas, respectively. Phase difference ranges from 9 min to 30 min in Barra and Caminha, respectively (Table 1). The highest differences between model predictions and observed data correspond to Caminha and Seixas stations, mainly due to bathymetric constraints that are more significant between Caminha and Vila Nova de Cerveira (Reis et al., 2009). The results for diurnal constituents reveal a good agreement between model predictions and observations, with average amplitude (phase) errors of about 10% (11.51) and 17% (10.61) for constituents K1 and O1, respectively. In summary, harmonic analysis results show that amplitude and phase of the major tidal constituents are well reproduced by the model. The water temperature and salinity predictions accuracy is also investigated. Salinity and water temperature data sampled near the bottom in Barra and Seixas stations during two tidal cycles was available for comparison with model predictions (Fig. 2). Although predicted (depth average values) and observed salinity/water temperature (bottom) are not referred to the same depth, this comparison shows that model reproduces the thermohaline properties variability, with small differences in temperature values. Maximum RMS was found for Barra station, with a value of 0.89 1C, which represents about 30% of the local water temperature amplitude. For salinity, RMS values are typically about 8% of the local amplitude. According to these results, it is considered that the model reproduces the heat and salt transport inside the Minho estuary and consequently was considered validated for the purposes of this research.

4.1.2. Rias Baixas The simulated period for the Rias Baixas model validation (L3, Fig. 1b) covers the period between November 1997 and May 1998 (the first six months of the simulation are considered as spin-up). This period was chosen considering the data availability and that during May 1998 was reported a high Minho River discharge (Alvarez et al., 2006), as well as favorable wind conditions to advect the estuarine plume toward the Rias Baixas, reversing the local estuarine circulation. Although these events are frequent and still occur presently they are not measured regularly in situ. The dataset used in the validation comprises harmonic constants of the main tidal constituents obtained at the mouth of the Rias Baixas (available from Puertos del Estado) and hydrographic parameters surveyed inside the Rias. To assess tidal model predictions accuracy for Rias Baixas (L3, Fig. 1b), SSE outputs from a hydrodynamic simulation from 1 to 18 May 1998 were compared with SSE computed through harmonic synthesis, using local harmonic constituents for Vigo (42114.40 N, 8143.80 W) and Villagarcia (42136.00 N, 8146.20 W) (Fig. 1b, black diamonds). RMS between predicted and computed time series is 0.06 m and 0.05 m for Villagarcia and Vigo stations, respectively. Predictive skills are close to 1 for both stations, confirming the excellent agreement between both datasets. Results of amplitude and phase for harmonic constants M2, S2, O1 and K1 determined from model predictions are presented in Table 2. The agreement between predicted and observed values is very good both in amplitude and in phase for semi-diurnal and diurnal constituents, which are the major tidal constituents in the Rias Baixas and near coastal region (Herrera et al., 2008; MartaAlmeida and Dubert, 2006). For M2 constituent, the difference between datasets is 0.01 m. In Villagarcia, the phase difference is 2.61, which means an average delay between observed and predicted tide of about 5.5 min for this constituent. For the Vigo station, the average delay is lower (about 2.5 min), revealing a good phase agreement. For the diurnal constituents, amplitude and phase agreement may be considered good for both stations

M.C. Sousa et al. / Continental Shelf Research 85 (2014) 30–41

35

Fig. 2. Observed and predicted salinity and water temperature time series for Barra and Seixas (Fig. 1c).

Table 2 Harmonic analysis results comparison of observed and predicted sea surface elevation data for Villagarcia and Vigo (M2, S2, O1 and K1 constituents). Tide gauge

M2 S2 O1 K1

Villagarcia Vigo Villagarcia Vigo Villagarcia Vigo Villagarcia Vigo

Amplitude (m)

Phase (1)

Data

Model

Difference

Data

Model

Difference

1.06 1.02 0.44 0.42 0.06 0.06 0.05 0.06

1.07 1.03 0.41 0.40 0.07 0.07 0.08 0.08


0.01
0.01 0.03 0.02
0.01
0.01
0.03
0.02

83.07 80.25 100.89 97.66 323.22 318.96 32.86 47.87

80.47 79.06 97.32 95.78 323.59 332.52 42.98 42.60

2.60 1.19 3.57 1.88
0.37
13.56
10.12 5.27

(about 0.01 m for amplitude and 101 for phase in Vigo). The results from the harmonic analysis show that tide should be classified as semidiurnal (form number E0.07) and that constituents M2 and S2 together determine about 90% of the astronomic tide in Western Galician coast. Following the aim of this research, the main topic of the validation is to assess model ability to reproduce the Minho estuarine plume dispersion pattern, which is analyzed comparing model predicted salinity vertical profiles with observations performed by Alvarez et al. (2006) inside the Rias Baixas (Fig. 3). The in situ salinity data available was measured weekly at three sampling stations located at the southern (wider and deeper) and northern mouths, and in the inner-middle areas of the Rias de Vigo, Pontevedra and Arousa during May 1998 (Fig. 1b). The in situ salinity data reveals an abnormal salinity gradient along the axis direction for the Rias de Vigo and Pontevedra. The southern mouths of the Rias are less saline than the innner region, i.e. an inverse circulation is observed. This difference is more marked in the Ria de Vigo. Reversely, the Ria de Arousa shows a different salinity gradient, being water near the mouth saltier than in the inner part of the estuary, corresponding to the typical pattern of an estuarine system. A similar situation is observed in model predictions. The salinity increase from south to north shows that low salinity values are not generated by the rivers inside the

rias, since the highest river runoff ( 200 m3 s
1) corresponds to the Ria de Arousa (the northernmost one) (Alvarez et al., 2006). Analysis of Fig. 3 shows that model overestimates salinity values in the southern mouth of the Ria de Vigo. Nonetheless, the halocline is well reproduced by the model for all stations, being predicted at the same depth. The maxima RMS values (1.27) between model predictions and measurements are observed in Ria de Vigo. For the other stations, RMS values range from 0.15 to 0.33. The bias was also determined, showing positive values for most stations, which indicate that model predictions tend to overestimate in situ salinity. The highest biases (about
1.19) are also observed in Ria de Vigo, and could be explained by an improper prescription of the landward model boundary condition. Model accuracy is also analyzed regarding the circulation pattern, using current meter data from the inner-middle part of Ria de Pontevedra (42123.510 N, 8144.290 W), measured on 12–13 May 1998 for a period of eight hours at six different depths during 5 min at each depth (Fig. 4). Both measured and predicted currents patterns show a negative estuarine circulation in the inner-middle part of Ria de Pontevedra (Fig. 4), which is characterized by near surface water moving landward and near bed water moving seaward, in accordance with previous findings presented by deCastro et al. (2004). In general, predictions reproduce the main features of observed velocity vertical structure, with small differences in intensity close to the surface. Observed maxima velocities are 0.10 m s
1 on the surface, while predicted velocities are 0.06 m s
1, meaning that model underestimates velocity in this region. No significant differences are observed near bed, showing a good agreement between predictions and measurements. It is important to note that, according to previous research in the area (deCastro et al., 2000), wind effects inside the estuary may dominate the current at surface layers, while bottom layers are mainly controlled by tide. Nevertheless, it should be considered that wind field used in this implementation (4 km resolution) does not have enough resolution to properly solve the main features inside the estuaries. Although errors are not negligible, especially on the surface layers, validation results show that the model developed in this study adequately reproduces the hydrodynamic behavior of the Rias Baixas, and in particular the intrusion of the Minho estuarine plume in these coastal regions.

36

M.C. Sousa et al. / Continental Shelf Research 85 (2014) 30–41

Fig. 3. Observed and predicted salinity vertical profiles for the sampling stations shown in Fig. 1b.

4.2. Negative circulation in Rias Baixas The model developed in this study was used to investigate the influence of the Minho estuarine plume in the generation of inverse circulation episodes in Rias Baixas, as the one described by Alvarez et al. (2006) in May 1998. Fig. 5 shows daily Minho River discharge (Fig. 5a, black line) and meridional wind component (Fig. 5b) at a point located close to the Minho River mouth (421N, 91W) for this period. River discharge shows an atypical pattern with high values during early May (1600 m3 s
1; Fig. 5a, black line). The meridional wind component is variable, with strong fluctuations in direction and intensity. The prevailing winds are from north, with intensities higher than 5 m s
1. Predicted surface salinity maps (Fig. 6, top) show a northward spread of the Minho River plume, which reaches the Ria de Vigo on 11 May. During the following days the northward displacement

of the plume continued and plume intrusion is observed in Ria de Pontevedra on 13–14 May. This situation generates an unusual surface salinity pattern at these locations. The along axis circulation in the mouths of Rias de Vigo, Pontevedra and Arousa is also calculated. Fig. 6a shows the circulation pattern at the Ria de Vigo calculated from 10 to 16 May. Between 11 and 13 May, an unusual circulation (upstream circulation) at the upper layers (up to 5–8 m) with water moving landward is observed, with highest velocities (0.30 m s
1) between 0 m and 2 m. Southwesterly winds during this period favored the reversal of positive circulation, resulting in introduction of the Minho estuarine outflow into the Ria. Chao (1988), Soares et al. (2007) and Marques et al. (2009), observed in their studies that downwelling winds reverses surface current, enhancing the mixing processes for La Plata River and Patos Lagoon. This reverse estuarine circulation may have some ecological consequences. It introduces the dinoflagellate

M.C. Sousa et al. / Continental Shelf Research 85 (2014) 30–41

37

Fig. 4. Observed (a) and predicted (b) along estuarine circulation (m s
1) in the inner-middle part of Ria de Pontevedra on 12–13 May 1998. Black line contours correspond to 0 m s
1.

blooms that are generated in coastal waters into the Rias (Fraga et al., 1988; Sordo et al., 2001), decreasing the abundance of marine species in this area. Thus, surface currents are mainly controlled by plume intrusion, while tidal effect is only observed at bottom layers. This pattern is in accordance with results obtained for Willapa Bay by Banas et al. (2004), where the Columbia River plume enters the mouth at all depths under a strong tidal flow and the axial gradient can in fact reverse for sustained periods. These forcing conditions confines the Minho plume close to the coast, where a less saline region is developed, as found by Otero et al. (2008). The Ria de Pontevedra (Fig. 6b) shows a similar circulation pattern although, in this case, surface water moving landward is observed between 13 and 15 May. In addition, this unusual circulation is more intense than for Ria de Vigo, with positive values until 10 m depth. In fact, around 13–14 May it is also observed bottom water moving seaward along one tidal cycle, which results in a negative estuarine circulation during this period. The results also agree with those reported by Alvarez et al. (2006), who found this negative estuarine circulation in a station located at the middle-inner estuary. Therefore, freshwater supplied by the Minho River could generate this unusual circulation, which tends to stop water exchange between this Ria and the shelf, increasing its residence time and consequently changing water quality. This reversal poses a selective force on the phytoplankton assemblage. Diatoms are unable to counteract

under these conditions and are therefore removed from the water column (Pitcher et al., 2010). On the other hand, high concentration of nutrients from the Minho River could fertilize the external part of the estuary, resulting in an extra feeding source for the main shellfish in the area (deCastro et al., 2006). Similar ecological consequences can be noticeable in estuaries located in upwelling regions (e.g. Pacific Northwest coast) (Hickey and Banas, 2003) or located north or south of the river mouth (Roegner et al., 2002). Finally, the Ria de Arousa (Fig. 6c) shows a normal estuarine circulation (outflow in surface layer and inflow in bottom layer) with the whole water column following the tidal cycle, landward during flood and seaward during ebb, without influence of the Minho estuarine plume. 4.3. Sensitive analysis of physical forcing on negative circulation River discharge and wind are important in the modulation of vertical and horizontal spreading of an estuarine plume. In fact, their temporal variability forces vertical and horizontal mix between buoyant and coastal waters. The presence or absence of the plume may provide an important environmental distinction between estuaries as well as between nearshore coastal regions. Thus, in order to assess the individual effect of Minho River outflow and wind forcing on the establishment of negative circulation in Rias Baixas, two new simulations were designed

38

M.C. Sousa et al. / Continental Shelf Research 85 (2014) 30–41

Fig. 5. Minho River discharge (a) and meridional wind component (b) between 19 April and 17 May 1998 (black line).

keeping most of the setups previously described (see model section). The first simulation considers essentially wind effect, adopting a constant river discharge value (833 m3 s
1) between 30 April and 4 May (Fig. 5a, dashed gray line). The second one analyses essentially river discharge effect, considering the real discharge (Fig. 5a, black line), but removing the wind from 1 to 13 May 1998. For the first scenario (removing the high river discharge from 1 to 3 May, Fig. 5a, dashed gray line), the obtained pattern is similar to that obtained in the simulation with real river discharge (Fig. 6a) with no significant differences at the Rias Baixas mouth. Choi and Wilkin (2007) also demonstrated through numerical modeling a high similarity between a steady-low and high discharge events in Hudson River mouth. Thus, only the along axis current for the Ria de Vigo is shown (Fig. 7a). These results show that the high Minho River discharge, observed at the beginning of May (Fig. 5a, dashed gray line), is not directly responsible for the abnormal hydrographic patterns found at the Rias Baixas, indicating that a continuous and moderate river discharge (500–1000 m3 s
1) may be enough to produce the negative circulation pattern. In fact, considering river runoff and wind data from 1979 to 2010 (period of available discharge for the Minho River), it were found 27% of events at moderate river discharge ( 4500 m3 s
1) under northward wind conditions, showing the importance of studying these situations. Without wind forcing (second scenario), the plume is displaced over the shelf, creating a bulge in front of the river mouth during first days (not shown). Then, the low salinity waters are advected to the right extending northward (along the coastline) reaching the Ria de Vigo mouth (Fig. 7b). In this situation, the plume effect

is observed earlier (between 6 and 8 May), affecting only near surface layers, being surface currents weaker in this case than in the real one (Fig. 6a), which is consistent with the results by Marques et al. (2009) for the Patos Lagoon. For the other Rias (not shown), the classical estuarine pattern is observed, showing that without wind forcing the Minho estuarine plume does not influence their circulation pattern. Note that the designed model application is unlikely to be unique to the Rias Baixas, suggesting that similar modeling approaches could be replicated to other coastal systems, such as Gulf of Mexico (Das et al., 2012), Atlantic coast of the United States (Guo and Valle-Levinson, 2007; Whitney and Garvine, 2005) or along the Washington–Oregon coast (Banas et al., 2004; Hickey and Banas, 2003), to improve understanding and characterize coupled estuarine-near coastal systems. Deeper knowledge of the influence of rivers plume on adjacent estuarine areas is a key issue, allowing the local scientist and managers to improve local policies of ecosystem protection.

5. Conclusions The main aim of this work is to study the propagation and influence of the Minho estuarine plume on Rias Baixas hydrography, and consequently to develop and explore an innovative numerical model application, integrating estuarine and coastal models. With this purpose, a nesting numerical model (MOHID) was implemented and validated for the Minho estuary and for the NW Iberian Coast, including the adjacent Rias Baixas. This integrated approach presents clear advantages in reproducing the realistic estuary-ocean interaction. Indeed, it takes into account

M.C. Sousa et al. / Continental Shelf Research 85 (2014) 30–41

39

Fig. 6. Evolution of surface salinity. Along estuarine circulation at stations located at mouths of Rias de Vigo (a), Pontevedra (b) and Arousa (c) between 10 and 15 May 1998. Black line contours correspond to 0 m s
1.

Fig. 7. Along estuarine circulation at station located at mouth of Ria de Vigo, considering constant Minho River discharge (a) and no wind forcing (b). Black line contours correspond to 0 m s
1.

40

M.C. Sousa et al. / Continental Shelf Research 85 (2014) 30–41

the estuarine dynamics (e.g. tide and ocean interaction) and not a river discharge represented by a climatological point source with constant salinity and water temperature data. Therefore, in this approach the salinity and water temperature horizontal gradients inside the estuary are computed considering the tidal propagation and the estuarine runoff is tidally modulated. Final results show that the adopted nesting methodology for the Minho estuary and Rias Baixas was successful implemented. Model predictions reproduce accurately the local hydrodynamics and thermohaline patterns. It was also concluded that under northward winds and Minho River discharges higher than 500 m3 s
1, the Minho River plume can reach Rias Baixas and reverse Rias de Vigo and Pontevedra circulation pattern, while the circulation of Ria de Arousa remains unchanged. The freshwater intrusion of the Minho River into these systems was the main responsible for the unusual horizontal salinity gradient (fresher water near the mouth than in the inner part) and negative circulation pattern characterized at these estuaries on previous research during May 1998. The influence of Minho River discharge and wind pattern in the event of May 1998 was also individually assessed considering two scenarios. It was concluded that the situation observed at the Rias Baixas at beginning of May was not directly dependent only on the high Minho River discharge observed. Without wind forcing, the Minho estuarine plume does not influence the circulation pattern in Rias Baixas. Furthermore, under northward winds, a continuous moderate Minho River discharge is enough to produce the negative circulation pattern, reducing the importance of specific events of high runoff values. This is an important feature, since, this negative circulation tends to stop water exchange between the Rias and shelf, increasing residence time and hence affecting water quality. An estuary that presents a large residence time is considered more sensitive to freshwater input because the water renewal is lower in these systems, affecting negatively its ecology. As Rias Baixas are recognized as an area of production of marine species of great economical interest, the existence of this negative circulation affects the exchange between the Rias and the ocean, changing the input of nutrients. Overall, the outcome of this study shows that the reverse circulation pattern observed in the Rias Baixas may be induced by a continuous moderate Minho River discharge under northward winds. Due to the high frequency of these moderate river discharge events (with 27% of occurrences) which spread the river plume toward the Rias Baixas under favorable wind conditions, the methodology proposed in this paper produces sound and thorough results, and proved to be useful and accurate enough to simulate the dynamics of the Minho estuarine plume along the Galician coast, as well as its effects on Rias Baixas. The necessary wind and river discharge conditions for the establishment of the negative circulation pattern in Rias Baixas will be researched in future studies.

Acknowledgements The first author of this work has been supported by the Portuguese Science Foundation through a doctoral grant (SFRH/ BD/60209/2009). The second author is supported by the Ciência 2008 program. The third author was supported through the Ramon y Cajal Program. This paper was partially supported by the Portuguese Science Foundation through the research projects DyEPlume (PTDC/MAR/107939/2008) and Pest (C/MAR/LA0017/ 2013) co-funded by COMPETE/QREN/UE, the Ministerio de Ciencia e Innovacion de España (Spain) through the GALINCLIMARCH Project (CGL2010-16688/BTE) and by Xunta de Galicia under the projects “Programa de Consolidación e Estructuración de Unidades

de Investigación Competitivas: Grupos de Referencia Competitiva” (GR2013-001) co-funded by European Regional Development Fund (FEDER) and EM2013/003. The authors acknowledge the Hydrographic Institute of the Portuguese Navy for providing sea surface elevation, salinity and water temperature data under the project ECOIS (POCTI/CTA/48461/2002).

References Alvarez, I., deCastro, M., Gomez-Gesteira, M., Prego, R., 2006. Hydrographic behavior of the Galician Rias Baixas (NW Spain) under the spring intrusion of the Mino River. J. Mar. Syst. 60, 144–152. Antonov, J.I., Seidov, D., Boyer, T.P., Locarnini, R.A., Mishonov, A.V., Garcia, H.E., Baranova, O.K., Zweng, M.M., Johnson, D.R., 2010. World Ocean Atlas 2009, volume 2: salinity. In: Levitus, S. (Ed.), NOAA Atlas NESDIS 69. U.S. Government Printing Office, Washington, DC, p. 184. Banas, N.S., Hickey, B.M., MacCready, P., Newton, J.A., 2004. Dynamics of Willapa Bay, Washington: a highly unsteady, partially mixed estuary. J. Phys. Oceanogr. 34, 2413–2427. Bettencourt, A., Ramos, L., Gomes, V., Dias, J.M.A., Ferreira, G., Silva, M., Costa, L., 2003. In: INAG (Ed.), Estuários Portugueses. Ordenamento do Território e Ambiente, Lisboa, p. 311. Blumberg, A.F., Kantha, L.H., 1985. Open boundary-condition for circulation models. J. Hydraul. Eng. ASCE 111, 237–255. Bruland, K.W., Lohan, M.C., Aguilar-Islas, A.M., Smith, G.J., Sohst, B., Baptista, A., 2008. Factors influencing the chemistry of the near-field Columbia River plume: nitrate, silicic acid, dissolved Fe, and dissolved Mn. J. Geophys. Res. Oceans, 113. Burchard, H., Petersen, O., Rippeth, T.P., 1998. Comparing the performance of the Mellor–Yamada and the k–epsilon two-equation turbulence models. J. Geophys. Res. Oceans 103, 10543–10554. Canuto, V.M., Howard, A., Cheng, Y., Dubovikov, M.S., 2001. Ocean turbulence. Part I: One-point closure model—momentum and heat vertical diffusivities. J. Phys. Oceanogr. 31, 1413–1426. Chao, S.Y., 1988. Wind-driven motion of estuarine plumes. J. Phys. Oceanogr. 18, 1144–1166. Chapra, S.C., 1997. Surface Water Quality Modeling. McGraw-Hill, New York, NY. Choi, B.J., Wilkin, J.L., 2007. The effect of wind on the dispersal of the Hudson River plume. J. Phys. Oceanogr. 37, 1878–1897. Das, A., Justic, D., Inoue, M., Hoda, A., Huang, H.S., Park, D.H., 2012. Impacts of Mississippi River diversions on salinity gradients in a deltaic Louisiana estuary: ecological and management implications. Estuarine Coastal Shelf Sci. 111, 17–26. deCastro, M., Alvarez, I., Varela, M., Prego, R., Gomez-Gesteira, M., 2006. Mino River dams discharge on neighbor Galician Rias Baixas (NW Iberian Peninsula): hydrological, chemical and biological changes in water column. Estuarine Coastal Shelf Sci. 70, 52–62. deCastro, M., Gomez-Gesteira, M., Alvarez, I., Prego, R., 2004. Negative estuarine circulation in the Ria of Pontevedra (NW Spain). Estuarine Coastal Shelf Sci. 60, 301–312. deCastro, M., Gomez-Gesteira, M., Prego, R., Taboada, J.J., Montero, P., Herbello, P., Perez-Villar, V., 2000. Wind and tidal influence on water circulation in a Galician ria (NW Spain). Estuarine Coastal Shelf Sci. 51, 161–176. Dias, J.M., Sousa, M.C., Bertin, X., Fortunato, A.B., Oliveira, A., 2009. Numerical modeling of the impact of the Ancao Inlet relocation (Ria Formosa, Portugal). Environ. Modell. Softw. 24, 711–725. Dortch, Q., Whitledge, T.E., 1992. Does nitrogen or silicon limit phytoplankton production in the Mississippi River plume and nearby regions. Cont. Shelf Res. 12, 1293–1309. Fiedler, P.C., Laurs, R.M., 1990. Variability of the Columbia River plume observed in visible and infrared satellite imagery. Int. J. Remote Sens. 11, 999–1010. Fong, D.A., Geyer, W.R., 2002. The alongshore transport of freshwater in a surfacetrapped river plume. J. Phys. Oceanogr. 32, 957–972. Fraga, F., Margalef, R., 1979. Las rias gallegas. Estudio y Explotacion del mar en Galicia. Santiago de Compostela, pp. 101–121. Fraga, S., Anderson, D.M., Bravo, I., Reguera, B., Steidinger, K.A., Yentsch, C.M., 1988. Influence of upwelling relaxation on dinoflagellates and shellfish toxicity in Ria-De-Vigo, Spain. Estuarine Coastal Shelf Sci. 27, 349–361. Freitas, V., Costa-Dias, S., Campos, J., Bio, A., Santos, P., Antunes, C., 2009. Patterns in abundance and distribution of juvenile flounder, Platichthys flesus, in Minho estuary (NW Iberian Peninsula). Aquat. Ecol. 43, 1143–1153. García-Berdeal, I., Hickey, B.M., Kawase, M., 2002. Influence of wind stress and ambient flow on a high discharge river plume. J. Geophys. Res. Oceans, 107. Garvine, R.W., 1995. A dynamical system for classifying buoyant coastal discharges. Cont. Shelf Res. 15, 1585. Gomez-Gesteira, M., Montero, P., Prego, R., Taboada, J.J., Leitao, P., Ruiz-Villarreal, M., Neves, R., Perez-Villar, V., 1999. A two-dimensional particle tracking model for pollution dispersion in A Coruna and Vigo Rias (NW Spain). Oceanol. Acta 22, 167–177. Guo, X.Y., Valle-Levinson, A., 2007. Tidal effects on estuarine circulation and outflow plume in the Chesapeake Bay. Cont. Shelf Res. 27, 20–42.

M.C. Sousa et al. / Continental Shelf Research 85 (2014) 30–41

Herrera, J.L., Varela, R.A., Roson, G., 2008. Spatial variability of the barotropic M (2) constituent tidal current over the Rias Baixas Galician shelf (NW Spain). J. Mar. Syst. 72, 189–199. Hickey, B.M., Banas, N.S., 2003. Oceanography of the US Pacific Northwest Coastal Ocean and estuaries with application to coastal ecology. Estuaries 26, 1010–1031. Hickey, B.M., Pietrafesa, L.J., Jay, D.A., Boicourt, W.C., 1998. The Columbia River plume study: subtidal variability in the velocity and salinity fields. J. Geophys. Res. Oceans 103, 10339–10368. Horner-Devine, A.R., Jay, D.A., Orton, P.M., Spahn, E.Y., 2009. A conceptual model of the strongly tidal Columbia River plume. J. Mar. Syst. 78, 460–475. IH, 2006. Tabela de Marés, vol. 1. Instituto Hidrográfico, Portugal 192. Jurisa, J.T., Chant, R., 2012. The coupled Hudson River estuarine-plume response to variable wind and river forcings. Ocean Dyn. 62, 771–784. Kortzinger, A., 2003. A significant CO2 sink in the tropical Atlantic Ocean associated with the Amazon River plume. Geophys. Res. Lett., 30. Le Provost, C., Lyard, F., Molines, J.M., Genco, M.L., Rabilloud, F., 1998. A hydrodynamic ocean tide model improved by assimilating a satellite altimeterderived data set. J. Geophys. Res. Oceans 103, 5513–5529. Leitão, P., 2002. Integração de Escalas e de Processos na Modelação do Ambiente Marinho. Instituto Superior Técnico 466. Leitão, P., Coelho, H., Santos, A., Neves, R., 2005. Modelling the main features of the Algarve coastal circulation during July 2004: a downscaling approach. J. Atmos. Ocean Sci. 10, 421–462. Lentz, S.J., Largier, J., 2006. The influence of wind forcing on the Chesapeake Bay buoyant coastal current. J. Phys. Oceanogr. 36, 1305–1316. Locarnini, R.A., Mishonov, A.V., Antonov, J.I., Boyer, T.P., Garcia, H.E., Baranova, O.K., Zweng, M.M., Johnson, D.R., 2010. World Ocean Atlas 2009, volume 1: temperature. In: Levitus, S. (Ed.), NOAA Atlas NESDIS 68. U.S. Government Printing Office, Washington, DC, p. 184. Lunven, M., Guillaud, J.F., Youenou, A., Crassous, M.P., Berric, R., Le Gall, E., Kerouel, R., Labry, C., Aminot, A., 2005. Nutrient and phytoplankton distribution in the Loire River plume (Bay of Biscay, France) resolved by a new Fine Scale Sampler. Estuarine Coastal Shelf Sci. 65, 94–108. Lyard, F., Lefevre, F., Letellier, T., Francis, O., 2006. Modelling the global ocean tides: modern insights from FES2004. Ocean Dyn. 56, 394–415. Marques, W.C., Fernandes, E.H., Monteiro, I.O., Moller, O.O., 2009. Numerical modeling of the Patos Lagoon coastal plume, Brazil. Cont. Shelf Res. 29, 556–571. Marta-Almeida, M., Dubert, J., 2006. The structure of tides in the Western Iberian region. Cont. Shelf Res. 26, 385–400. Martinsen, E.A., Engedahl, H., 1987. Implementation and testing of a lateral boundary scheme as an open boundary-condition in a barotropic ocean model. Coastal Eng. 11, 603–627. Monteiro, I.O., Marques, W.C., Fernandes, E.H., Goncalves, R.C., Moller, O.O., 2011. On the effect of earth rotation, river discharge, tidal oscillations, and wind in the dynamics of the Patos Lagoon coastal plume. J Coastal Res 27, 120–130. Monteiro, P.M.S., Largier, J.L., 1999. Thermal stratification in Saldanha Bay (South Africa) and subtidal, density-driven exchange with the coastal waters of the Benguela upwelling system. Estuarine Coastal Shelf Sci. 49, 877–890. Mourino, C., Fraga, F., 1982. Hydrography of the Ria of Vigo – 1976–1977 – abnormal influence of the Mino River. Invest. Pesq. 46, 459–468. Otero, P., Ruiz-Villarreal, M., García-García, L., González-Nuevo, G., Cabanas, J.M., 2013. Coastal dynamics off Northwest Iberia during a stormy winter period. Ocean Dyn. 63, 115–129. Otero, P., Ruiz-Villarreal, M., Peliz, A., 2008. Variability of river plumes off Northwest Iberia in response to wind events. J. Mar. Syst. 72, 238–255. Otero, P., Ruiz-Villarreal, M., Peliz, A., Cabanas, J.M., 2010. Climatology and reconstruction of runoff time series in northwest Iberia: influence in the shelf buoyancy budget off Ria de Vigo. Sci. Mar. 74, 247–266. Ouellet, V., Secretan, Y., Hilaire, A., Morin, J., 2013. Daily Averaged 2D Water Temperature Model for the St. Laurence River. River Research and Applications John Wiley & Sons, Ltd., pp. 1–12. Pearcy, W.G., 1992. Ocean Ecology of North Pacific Salmonids. Pitcher, G.C., Figueiras, F.G., Hickey, B.M., Moita, M.T., 2010. The physical oceanography of upwelling systems and the development of harmful algal blooms. Prog. Oceanogr. 85, 5–32.

41

Plus, M., Dumas, F., Stanisiere, J.Y., Maurer, D., 2009. Hydrodynamic characterization of the Arcachon Bay, using model-derived descriptors. Cont. Shelf Res. 29, 1008–1013. Prego, R., Fraga, F., 1992. A simple-model to calculate the residual flows in a Spanish Ria—hydrographic consequences in the Ria-De-Vigo. Estuarine Coastal Shelf Sci. 34, 603–615. Reifel, K.M., Johnson, S.C., DiGiacomo, P.M., Mengel, M.J., Nezlin, N.P., Warrick, J.A., Jones, B.H., 2009. Impacts of stormwater runoff in the Southern California Bight: relationships among plume constituents. Cont. Shelf Res. 29, 1821–1835. Reis, J.L., Martinho, A.S., Pires-Silva, A.A., Silva, A.J., 2009. Assessing the influence of the river discharge on the Minho Estuary tidal regime. J. Coastal Res., 1405–1409. Roegner, G.C., Hickey, B.M., Newton, J.A., Shanks, A.L., Armstrong, D.A., 2002. Windinduced plume and bloom intrusions into Willapa Bay, Washington. Limnol. Oceanogr. 47, 1033–1042. Ruiz-Villarreal, M., Montero, P., Taboada, J.J., Prego, R., Leitao, P.C., Perez-Villar, V., 2002. Hydrodynamic model study of the Ria de Pontevedra under estuarine conditions. Estuarine Coastal Shelf Sci. 54, 101–113. Santoro, P., Fernandez, M., Fossati, M., Cazes, G., Terra, R., Piedra-Cueva, I., 2011. Preoperational forecasting of sea level height for the Rio de la Plata. Appl. Math. Modell. 35, 2462–2478. Soares, I.D., Kourafalou, V., Lee, T.N., 2007. Circulation on the western South Atlantic continental shelf: 2. Spring and autumn realistic simulations. J .Geophys. Res. Oceans, 112. Sordo, I., Barton, E.D., Cotos, J.M., Pazos, Y., 2001. An inshore poleward current in the NW of the Iberian peninsula detected from satellite images, and its relation with G-catenatum and D-acuminata blooms in the Galician Rias. Estuarine Coastal Shelf Sci. 53, 787–799. Sousa, M.C., Alvarez, I., Vaz, N., Dias, J.M., 2011. Physical forcing of the hydrography of the Ria de Vigo mouth. J. Coastal Res. SI64, 1589–1593. Sousa, M.C., Dias, J.M., 2007. Hydrodynamic model calibration for a mesotidal lagoon: the case of Ria de Aveiro (Portugal). J. Coastal Res. SI50, 1075–1080. Sousa, R., Guilhermino, L., Antunes, C., 2005. Molluscan fauna in the freshwater tidal area of the River Minho estuary, NW of Iberian Peninsula. Ann. Limnol. Int. J. Limnol. 41, 141–147. Swarzenski, C.M., Doyle, T.W., Fry, B., Hargis, T.G., 2008. Biogeochemical response of organic-rich freshwater marshes in the Louisiana delta plain to chronic river water influx. Biogeochemistry 90, 49–63. Taboada, J.J., Prego, R., Ruiz-Villarreal, M., Gomez-Gesteira, M., Montero, P., Santos, A.P., Perez-Villar, V., 1998. Evaluation of the seasonal variations in the residual circulation in the Ria of Vigo (NW Spain) by means of a 3D baroclinic model. Estuarine Coastal Shelf Sci. 47, 661–670. Tenore, K.R., Alonsonoval, M., Alvarezossorio, M., Atkinson, L.P., Cabanas, J.M., Cal, R.M., Campos, H.J., Castillejo, F., Chesney, E.J., Gonzalez, N., Hanson, R.B., Mcclain, C.R., Miranda, A., Roman, M.R., Sanchez, J., Santiago, G., Valdes, L., Varela, M., Yoder, J., 1995. Fisheries and oceanography off Galicia, NW Spain— mesoscale spatial and temporal changes in physical processes and resultant patterns of biological productivity. J. Geophys. Res. Oceans 100, 10943–10966. Varela, R.A., Roson, G., Herrera, J.L., Torres-Lopez, S., Fernandez-Romero, A., 2005. A general view of the hydrographic and dynamical patterns of the Rias Baixas adjacent sea area. J. Mar. Syst. 54, 97–113. Vaz, N., Dias, J.M., Leitao, P.C., 2009a. Three-dimensional modelling of a tidal channel: the Espinheiro Channel (Portugal). Cont. Shelf Res. 29, 29–41. Vaz, N., Fernandes, L., Leitao, P.C., Dias, J.M., Neves, R., 2009b. The tagus estuarine plume induced by wind and river runoff: winter 2007 case study. J. Coastal Res. SI56, 1090–1094. Warner, J.C., Geyer, W.R., Lerczak, J.A., 2005. Numerical modeling of an estuary: a comprehensive skill assessment. J. Geophys. Res. Oceans, 110. Warrick, J.A., Milliman, J.D., 2003. Hyperpycnal sediment discharge from semiarid southern California rivers: implications for coastal sediment budgets. Geology 31, 781–784. Whitney, M.M., Garvine, R.W., 2005. Wind influence on a coastal buoyant outflow. J. Geophys. Res. Oceans, 110. Wong, K.C., Lu, X., 1994. Low-frequency variability in Delaware Inland Bays. J. Geophys. Res. Oceans 99, 12683–12695. Yankovsky, A.E., Chapman, D.C., 1997. A simple theory for the fate of buoyant coastal discharges. J. Phys. Oceanogr. 27, 1386–1401.