NW Iberia shelf dynamics: The River Douro plume

© 3rd

IAHR Europe Congress, Book of Proceedings, 2014, Porto - Portugal. ISBN 978-989-96479-2-3

NW IBERIA SHELF DYNAMICS: THE RIVER DOURO PLUME

(1) _{Centro Interdisciplinar de Investigação Marinha e Ambiental, Universidade do Porto, Porto, Portugal} [email protected], [email protected]

(2) _{Instituto de Hidráulica e Recursos Hídricos, Universidade do Porto, Porto, Portugal} (3) _{Centro de Ciências Matemáticas, Universidade da Madeira, Funchal, Portugal}

[email protected]

(4) _{Faculdade de Engenharia, Universidade do Porto, Porto, Portugal} [email protected]

Abstract

The River Douro, one of the major rivers of the Iberian Peninsula, is located on the NW Iberian coast. Its daily averaged discharge can range from 0 to 13 000 m3_{/s, which impacts on the} formation of the river plume and on its dispersion along the continental shelf. The scenarios of plume generation, retention and dispersion were simulated with the ROMS model. The results show that the plumes behaviour depends on the wind strength and direction and on the offshore geostrophic current system. It was also found that the plume might induce offshore mesoscale eddies that transport the plume waters across-shore.

Keywords: Cross-shore transport, alongshore currents, advection, buoyant plumes.

1. Introduction

One of the most important mechanisms that transport terrestrial materials to the ocean is the river plumes. Rivers carry essential nutrients, enhancing the phytoplankton productivity, and sediments, which settle on the seabed producing modifications on the bathymetry that might affect the navigation channels. The study and modelling of the river plumes is a key factor for the complete understanding of sediment transport mechanisms and patterns, and of coastal physics and dynamic processes.

The three main processes that govern plume dynamics are the mixing induced by the turbulence, the alongshore current produced by the balance between the Coriolis force and the cross-shore pressure gradient and the acceleration produced by the balance between buoyancy and gravity forces (McClimans, 1986). Plumes are superficial structures that can be trapped in the inner shelves with a baroclinic boundary coastal current structure (Chao and Boicourt, 1986) but they occupy a small portion of the total water column (Münchow and Garvine, 1993) and their behaviour varies with the wind characteristics, tides, bottom friction on shallow waters, and offshore currents. When the riverine waters reach the ocean, instabilities can be induced, and, on this mixing, several structures can be formed like, for instance, bulges, filaments, buoyant currents, fronts, current-jets, eddies, etc.

River plumes have been the subject of numerical modelling for the last two decades. Two of the first works on river plumes with simplified 3D primitive equations models were those of Chao and Boicourt (1986) and Chao (1988), which established a plume classification still in use. This classification distinguishes between subcritical and supercritical plumes according to the size of the two principal regions of a plume: bulge and coastal current. If the seaward extent of the bulge is comparable to the width of the coastal current the plume is subcritical. Otherwise, if the bulge presents a considerable seaward extent, for a narrow alongshore current, the plume is supercritical.

The River Douro is one of the major rivers of the Iberian Peninsula. Its daily averaged flow rate, controlled 21.6 km-upstream from its mouth by the Crestuma dam, ranges from 0 to 13000 m3_{/s. The river outflow presents an important inter-annual variability, and a normal} annual variability, with strong fluxes in winter and weak ones in summer (see Table 1). About 70 % of the annual precipitation in this region take place between October and May with the highest values between December and February, and lowers between July and September (Gómez-Gesteira et al., 2011). The weather characteristics are mediated by the Azores Anticyclone (Lorenzo et al., 2007), which produce winds with northerly and north-westerly directions (Ramos et al., 2011) more frequent in summer. These northern-direction winds produced upwelling events, where a narrow band of cold batter is formed along the coast (Alvarez et al., 2008; deCastro et al., 2008). The southerly, south-westerly and westerly winds are at their highest amplitude during the autumn and winter months.

Table 1. River Douro mean temperature and monthly averaged discharge.

MONTH Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

T(ºC) 7 8 9 10 12 15 18 20 18 14 12 10

Q(m3_{/s) 1083 855 762 522 402 272 165 101 160 273 473 862}

The NW Iberian coast, where the Douro plume evolves, presents prominent capes, promontories and submarine canyons (see Figure 1) with a continental shelf 30–40 km-wide between Galicia and Aveiro, 50–55 wide between Aveiro and Lisbon, and less than 20 km-wide south of Lisbon (Pinheiro et al., 1996; Mason et al., 2006; Relvas et al., 2007). The semidiurnal tides present an amplitude between 2 and 3.4 m (Benavent et al., 2009), with a small amplification along the northern Portuguese shelf and a south–north propagation of the tidal wave as a Kelvin wave with a phase lag of 1 h between the south and the north of Portugal (Marta-Almeida and Dubert, 2006; Sauvaget et al., 2000). The effect of all the sources of fresh water on the Western Iberian coast (the Galician Rias, and the rivers Minho, Mondego, Lima, Vouga and Tejo) produces the Western Iberian Buoyant Plume (Peliz et al., 2002), extending along the coast and forming a front with the warmer and more saline surface waters. It increases the stratification and produces a vertical retention keeping the plume materials (biological materials, sediments, pollutants, etc.) inshore (Martins et al., 2008; Relvas et al., 2007; Silva et al., 2008; Otero et al., 2009).

Figure 1. Model domain limits and bathymetry (m).

The aim of this work is to study the River Douro plume and its dynamic on the inner- and cross-shelf regions, building on the long-term objective of generate a Douro River plume forecasting system as part of the RAIA and RAIA.co projects.

2. Material and methods

The Regional Oceanic Modeling System (ROMS) (Shchepetkin and McWilliams, 2003, 2005) was selected to perform the study of the Douro river plumes. It is a split-explicit, free-surface and terrain-following vertical coordinate oceanic model. A larger time-step is used for the temperature and salinity transport equations and baroclinic momentum equation, while a smaller time-step is used for the barotropic momentum and continuity equations. It contains a two-way procedure for the barotropic mode satisficing the 3D continuity equation, and a third-order, upwind, dissipative advection scheme for momentum allowing for steep gradients and thus enhancing the resolution of the solution for a given grid size (Shchepetkin and McWilliams, 1998). The atmospheric forcing, the initial and boundary conditions and the model grid were constructed using the ROMS TOOLS package (Penven et al., 2007). The bottom topography was extracted from the GEBCO 30 arc-second grid resolution database (GEBCO, 2010). The initial conditions and forcing at the oceanic boundaries, temperature and salinity were retrieved from the World Ocean Atlas climatology (Antonov et al., 2006; Locarnini et al., 2006), the atmospheric fluxes, water and heat, came from COADS (da Silva et al., 1994), while tidal forcing resulted from the OSU tidal data inversion (Egbert and Erofeeva, 2002). The computational domain has 35 vertical levels with stretched S-coordinates (Song and Haidvogel, 1994), thus ensuring an acceptable resolution of the upper layer, and extends on the Earth’s surface from 10.75° W to 8° W and from 39.5° N to 42.5° N (see Figure 1), on a 1°/80 (∼1 km) regular grid.

A 3 year spin-up simulation without river flux is needed for the volume averaged kinetic energy to reach a steady-state condition. After this 3 year-long simulation, a 1 month-long tide-ramp simulation was performed. The outcome of this procedure is to be used as initial conditions for the case-study simulations, where the river flow is added like a point source located at the River Douro location with a specified temperature and null salinity.

A monthly mean flow rate at the Crestuma dam, averaged for the Jan/1996–Jan/2012 period (SNIRH, n.d.) was used. For analysis, the computed salinity and current velocity distributions were used to represent the orientation, structure and propagation of the formed plumes. A plume classification based on phenomenological parameters was performed. Chao (1988) classifies the plumes according to the ratio between the maximum width of the plume bulge and the coastal current width. For a ratio bellow 1.7, the plume is subcritical, otherwise it is supercritical. The Rossby deformation radius, Rd, could be used to quantify the planetary rotation effect on the plume structure. This number is defined as:

N h

 Rd

,

f

=

where, h is the local ocean depth scale, and f is the Coriolis parameter given by:

f

= Ω

2

sin ,

φ

with Ω = 7.29 x 10-5 _{rad/s being Earth’s rotational velocity), and} φ _{the local latitude. N is the} Brunt-Väisälä frequency calculated from:

g

N

.

z

ρ

ρ

∂

= −

∂

In Eq [3], g is the acceleration of gravity, ρθ is the monthly mean potential density and z is the geometric height.

The Kelvin number, Ke, is the ratio between the local plume width and the Rossby deformation radius. For Ke << 1, the Coriolis force will have a weak effect over the plume structure and the advection due to momentum will be strong. If Ke >> 1 the situation is the opposite: strong Coriolis force effect and weak advection (Garvine, 1995, 1999).

The densimetric Richardson number, Ri can be defined as: 2

g 'h

 Ri

,

U

=

where U is the horizontal depth averaged velocity, and g’ is the reduced gravity, given by

b s

 g ' g

ρ

ρ

.

ρ

−

=

In Eq [5], ρb and ρs are the bottom and surface water densities, and ρ is the depth averaged water density. This number characterises the water column stability. Ri = 20 is the upper limit for occurrence of turbulent mixing near the halocline. For Ri < 20 a weak stratification situation exists, while for Ri > 20 the mixing is negligible. For Ri < 2 the mixing will be fully developed (Schiller and Kourafalou, 2010; Vaz et al., 2012)

Three types of simulations were performed: 1. schematic winds simulations with prescribed river outflow, and wind speed and direction; 2. a multi-year climatological simulation, with monthly prescribed river outflow and water temperature; 3. extreme case simulation, based on the Entre-os-Rios accident situation.

3. Results

3.1 Schematic wind simulations

The 1 month-long simulations were performed with January climatic initial conditions at the open ocean lateral boundaries. The wind and river flow conditions for each scenario are summarized on Table 2.

Table 2. Schematic wind scenarios performed with the ROMS model. Scenario WIND DIRECTION WIND VELOCITY RIVER FLOW

(m/s) (m3_/s) A1 North 12 1000 A2 South 12 1000 B1 North 1 1000 B2 South 1 1000 C1 North 12 100 C2 South 12 100 D1 North 1 100 D2 South 1 100 E1 No wind No wind 1000 E2 No wind No wind 100

The main results can be seen on Figure 2. For the no wind cases, the plumes developed weak poleward travelling coastal currents and bulges. Comparing scenarios E1 and E2 with the other cases, a strong wind response can be related with river plumes propagating with the wind direction near the coast. Weak wind simulations, i.e. scenarios B1, B2, D1 and D2, produce the larger plume bulge widths, which can cross the continental shelf, in the case of a simultaneous strong river outflow, as in scenarios B1 and B2. For a strong wind, scenarios A1, A2, C1 and C2, the bulges are small or even non-existent. The southerly winds generate poleward travelling plumes which might reach the Galician Rias when a strong river flow occurs. The northerly winds produce plumes with south-westerly direction and south alongshore currents. As can be seen on Table 3, the maximum plume widths are obtained for strong river outflows.

Table 3. Chaos’s plume classification.

Scenario BULGE WIDTH (km)

COASTAL CURRENT WIDTH (km)

CHAO’S RATIO

Maximum River mouth

A1 45 5 15 3.0 Supercritical A2 — 5 10 0.1 Subcritical B1 48 37 14 3.4 Supercritical B2 32 30 15 2.1 Supercritical C1 — 3 5 0.1 Subcritical C2 — 3 10 0.1 Subcritical D1 24 17 5 4.8 Supercritical D2 24 12 10 2.4 Supercritical E1 30 24 10 3.0 Supercritical E2 12 8 5 2.4 Supercritical EE 33 30 15 2.2 Supercritical

A1 A2 B1 B2

C1 C2 D1 D2

E1 E2 EE

Figure 2. Mean salinity and current vectors for the one month schematic wind scenarios simulations, A1 to E2, and for the extreme event study, EE.

3.2 Extreme event study

On the 4th _{March 2001, the river Douro flow rate reached 7000 to 8000 m}3_{/s. This extremely} strong discharge provoked the collapse of a bridge at Entre-os-Rios, and one bus and several were carried downstream. A few days after the accident, some corpses and car debris were found on the Galician coast. For the simulation, QuikSCAT (Dunbar et al., 2006) real wind data, and a mean river flow of 8000 m3_{/s were used. River water salinity and temperature were set} to null and 9 ºC, respectively. The results are shown in Figure 2. The plume displays a huge bulge and a northward coastal current that can explain the corpses and car debris on the Galician coast. The wind presents a southerly wind component, driving the plume waters to the north.

3.3 Multi-year climatological simulation

On this 3 year-long simulation the river water salinity was kept null and the temperature and flow rate change according to the values shown in Table 1. The temperature values were extracted from previous works (Azevedo et al., 2006; Pinto, 2007). Figure 3 shows the variation of the plume structure with the climatological conditions in the area. On January, a southwest protruding jet-like plume can be seen, while on December the bulge and the coastal current seem to evolve to the north. On May the bulge evolved towards southwest but a small northward component is present, and on August, due to the small river flow, the bulge is small. Although this fact concurs with the climatic wind conditions, the analysis of the daily snapshots shows that the Douro plume is reactive to the offshore (alongshore) current and there are times when the wind and the plume propagate in opposite directions. This happens when the offshore current is at its highest strength and travels opposite to the wind. This simulation displays a wide range of plume scenarios: jet-like features, bifurcations, jets protruding either to north or to south, patches, flapping filaments and mesoscale eddies. These structures interact with inshore waters and advect the plume waters across the shelf transporting them to remote locations.

January May August December

Figure 3. Monthly mean salinity and current vectors for four months of the multi-year climatological simulation.

3.4 Plume classification

This Chao’s classification was applied to the 1 month-long simulations and to the extreme event study (EE). This led to the classification of 3 plumes as subcritical, and 8 plumes as supercritical (see Table 3).

Deformation radius ranging from 5 to 10 km, and reaching up to 18 km for the EE, show that the Douro plumes are not diffusive. The Kelvin number is close to 0 for the subcritical plumes and between 1.8 and 5.6 for the supercritical ones, which shows that the subcritical plumes have a stronger effect of advection and that the supercritical plumes have an important influence of the Coriolis force.

The largest Richardson numbers, Ri > 20, were registered for low or absent wind and visible bulges, and for the EE as well. This means that there was no efficient mixing in those cases. Scenarios C1 and C2 present Ri < 2, suggesting strong mixing.

4. Conclusions

The simulations shown the typical structure of a river plume, i.e. bulge plus coastal current. The schematic wind case-studies suggest that the plume of the Douro is wind-driven, showing remarkable differences in their structure and dispersion pathways depending on the wind strength and direction. Southerly winds push the river water to the north, being possible to find water from the River Douro within the Galician Rias, while upwelling favourable winds induce plumes with a narrow coastal current. The high surface salinity on the plume regions during strong wind events suggests that the wind enhances the vertical mixing. The plume is affected by the Coriolis force but its influence is mitigated by a strong wind forcing.

The multi-year climatological study showed that the plume response depends as well on the behaviour of the offshore geostrophic current system. Offshore eddies and filaments are also responsible for the cross-shore transport, through the horizontal advection of plume waters. Extreme river discharges, associated with southerly winds, can transport debris to the Galician coast, helping to explain the outcome of the tragic events of the Entre-os-Rios accident of March 2001.

Analysis of both the Rossby deformation radius and the Kelvin number confirm that the Douro supercritical plumes are strongly affected by the planetary rotation. The supercritical plumes coincided with the coastal current maximum widths and the values obtained for the densimetric Richardson number showed that the supercritical plumes are less mixed than the subcritical ones.

Acknowledgments

Isabel Iglesias wants to thanks RAIA (0313-RAIA-1-E) and RAIA.co (0520-RAIA-CO-1-E) projects that provided her postdoctoral funds. Rui Caldeira was supported by funds from the ECORISK project (NORTE-07-0124-FEDER-000054). The authors would also like to thank Luísa Bastos, Isabel Gonçalves and Iria Sala for the fruitful discussions, and the directors of the IHRH, Prof. Veloso-Gomes and Prof. Taveira-Pinto for the provided logistic support. The RAIA Coastal Observatory has been funded by the Programa Operativo de Cooperación Transfronteriza España–Portugal (POCTEP 2007–2013).

References

Álvarez I, Gómez-Gesteira M, deCastro M, Dias M 2008. 'Spatiotemporal evolution of upwelling regime along the western coast of the Iberian Peninsula', Journal of Geophysical

Research: Oceans, 113, C07020.

Antonov JI, Locarnini RA, Boyer TP, Mishonov AV, Garcia HE 2006. 'World Ocean Atlas 2005. Volume 2: Salinity'. Volume 62 of NOAA Atlas NESDIS. U.S. Department of Commerce, NOAA, NESDIS, Washington, D.C.

Azevedo IC, Duarte PM, Bordalo AA 2006. 'Pelagic metabolism of the Douro estuary (Portugal) — Factors controlling primary production'. Estuarine, Coastal and Shelf Science, 69, 133–146. Benavent M, Arnoso J, Montesinos FG 2009. 'Regional ocean tide loading modelling around the

Iberian Peninsula', Journal of Geodynamics, 48, 132–137.

Chao S 1988. 'River-forced estuarine plumes', Journal of Physical Oceanography, 18, 72–88. Chao S, Boicourt WC 1986. 'Onset of estuarine plumes', Journal of Physical Oceanography, 16,

daSilva AM, Young-Molling CC, Levitus S 1994. Atlas of Surface Marine Data 1994. Volume 1: Algorithms and Procedures. Volume 6 of NOAA Atlas NESDIS. U.S. Department of Commerce, NOAA, NESDIS, Washington, D.C.

deCastro M, Gómez-Gesteira M, Lorenzo MN, Álvarez I, Crespo AJC 2008. 'Influence of atmospheric modes on coastal upwelling along the western coast of the Iberian Peninsula, 1985 to 2005', Climate Research, 36, 169–179.

Dunbar RS, Lungu T, Weiss B, Stiles B, Huddleston J, Callahan PS, Shirtliffe G, Perry KL, Hsu C, Mears C, Wentz F, Smith D 2006. QuikSCAT Science Data Product User Manual, Version 3.0, JPL Document D–18053 - Rev A. Jet Propulsion Laboratory. Pasadena, CA.

Egbert GD, Erofeeva SY 2002. 'Efficient inverse modeling of barotropic ocean tides', Journal of

Atmospheric and Ocean Technology, 19, 183–204.

Garvine RW 1995. 'A dynamical system for classifying buoyant coastal discharges', Continental

Shelf Research, 15, 1585–1596.

Garvine RW 1999. 'Penetration of buoyant coastal discharge onto the continental shelf: A numerical model experiment', Journal of Physical Oceanography, 29, 1892–1909.

GEBCO 2010. The GEBCO_08 Grid, version 20100927, http://www.gebco.net.

Gómez-Gesteira M, Gimeno L, deCastro M, Lorenzo MN, Alvarez I, Nieto R, Taboada JJ, Crespo AJC, Ramos AM, Iglesias I, Gómez-Gesteira JL, Santo FE, Barriopedro D, Trigo IF 2011. 'The state of climate in NW Iberia', Climate Research, 48, 109–144.

Locarnini RA, Mishonov AV, Antonov JI, Boyer TP, Garcia HE 2006. 'World Ocean Atlas 2005. Volume 1: Temperature'. Volume 61 of NOAA Atlas NESDIS. U.S. Department of Commerce, NOAA, NESDIS, Washington, D.C.

Lorenzo MN, Taboada JJ, Gimeno L 2007. 'Links between circulation weather types and teleconnection patterns and their influence on precipitation patterns in Galicia (NW Spain)',

International Journal of Climatology, 28, 1493–1505.

Marta-Almeida M, Dubert J 2006. 'The structure of tides in the Western Iberian region',

Continental Shelf Research, 26, 385–400.

Martins I, Silva A, Santos A, Bastos L 2008. 'Assessing critical runoff values for the generation of a buoyancy induced coastal current off the western Iberian Peninsula', in: Geophysical Research Abstracts, European Geosciences Union.

Mason E, Coombs S, Oliveira PB 2006. 'An overview of the literature concerning the oceanography of the eastern North Atlantic region', Relatórios Técnicos e Científicos, Série Digital 33, IPIMAR, Lisboa, Portugal.

McClimans TA 1986. 'Estuarine fronts and river plumes', Dronkers J, van Leussen W (Eds.), Physical Processes in Estuaries, Springer-Verlag, New York, 55–69. ISBN 9783540192664 Münchow A, Garvine RW 1993. 'Dynamical properties of a buoyancy driven coastal current',

Journal of Geophysical Research: Oceans, 98, 20063–20077.

Otero P, Ruiz-Villarreal M, Peliz Á 2009. 'River plume fronts off NW Iberia from satellite observations and model data', ICES Journal of Marine Science, 66, 1853–1864.

Peliz Á, Rosa TL, Santos AMP, Pissarra JL 2002. 'Fronts, jets, and counter-flows in the Western Iberian upwelling system', Journal of Marine Systems, 35, 61–77.

Penven P, Marchesiello P, Debreu L, Lefèvre J 2007. 'Software tools for pre- and post-processing of oceanic regional simulations', Environmental Modelling & Software, 23, 660–662. Pinheiro LM, Wilson RCL, Pena dos Reis R, Whitmarsh RB, Ribeiro A 1996. 'The western Iberia margin: A geophysical and geological overview', Whitmarsh RB, Sawyer DS, Klaus A, Masson DG (Eds.), Proceedings of the Ocean Drilling Program, College Station, TX (Ocean Drilling Program), 3–23.

Pinto J 2007. 'Influência do Regime de Escoamento Fluvial na Hidrologia e Dinâmica do Estuário do Douro'. Relatório Final de Estágio. Universidade de Évora.

Ramos AM, Ramos R, Sousa P, Trigo RM, Janeira M, Prior V, 2011. ' Cloud to ground lightning activity over Portugal and its association with circulation weather types', Atmospheric

Research, 101, 84–101.

Relvas P, Barton ED, Dubert J, Oliveira PB, Peliz Á, daSilva JCB, Santos AMP 2007. 'Physical oceanography of the western Iberia ecosystem: Latest views and challenges', Progress in

Oceanography, 74, 149–173.

Sauvaget P, David E, Guedes Soares C 2000. 'Modelling tidal currents on the coast of Portugal',

Coastal Engineering, 40, 393–409.

Schiller RV, Kourafalou VH 2010. 'Modeling river plume dynamics with the HYbrid Coordinate Ocean Model', Ocean Modelling, 33, 101–117.

Shchepetkin AF, McWilliams JC 1998. 'Quasi-monotone advection schemes based on explicit locally adaptive dissipation', Monthly Weather Review, 126, 1541–1580.

Shchepetkin AF, McWilliams JC 2003. 'A method for computing horizontal pressure-gradient force in an oceanic model with a nonaligned vertical coordinate', Journal of Geophysical

Research: Oceans, 108, 3090.

Shchepetkin AF, McWilliams JC 2005. 'The regional oceanic modeling system (ROMS): A split-explicit, free-surface, topography-following coordinate oceanic model', Ocean Modelling, 9, 347–404.

Silva A, Martins L, Bastos L 2008. 'An intermittent coastal current generated by buoyancy and wind forcing', in: Geophysical Research Abstracts, European Geosciences Union.

SNIRH, n.d. Serviço Nacional de Informação de Recursos Hídricos [WWW]. Available from: http://snirh.pt [Accessed 08/Jan/2013].

Song Y, Haidvogel D 1994. 'A semi-implicit ocean circulation model using a generalized topography-following coordinate system', Journal of Computational Physics, 115, 228–244. Vaz N, Lencart e Silva JD, Dias JM 2012. 'Salt fluxes in a complex river mouth system of