Modeling Stratification Mixing Processes at the mouth of a Dam Controlled River

Transcription

1 European Water 13/14: 1-8, 6. 6 E.W. Publications Modeling Stratification Mixing Processes at te mout of a Dam Controlled River G. Sylaios 1,, V. Tsirintzis and K. Haralambidou 1 National Agricultural Researc Foundation, Fiseries Researc Institute, Nea Peramos 64 7, Kavala, Greece. Laboratory of Ecological Engineering and Tecnology, Department of Environmental Engineering, Scool of Engineering Democritus University of Trace, 671 Xanti, Greece. Abstract: Key words: In tis paper te question of freswater induced stratification at te mout of a dam-controlled river is addressed. Understanding te development and breakdown of stratification is a key environmental issue for river mouts, since it largely determines te vertical fluxes of water properties suc as eat, salt, momentum and nutrients. An intensive monitoring program (vertical profiles of salinity, temperature and water flow) as been undertaken along Strymon river estuary during te spring and summer period, wen river flow is man-controlled from Kerkini dam, according to irrigation needs. Te main target of tis monitoring program was to provide information on te stratification mixing conditions prevailing in te area of Strymon River mout under different water flow regimes. A stratification mixing model in terms of te water column potential energy (φ) was developed and applied in Strymon River, to account separately for te local contributions to stratification (solar eating and freswater buoyancy) and mixing (wind and bottom sear stresses). Te model was run under two different scenarios (te influence of Kerkini dam and te absence of Kerkini dam) to assess te impact of water retention on te stratification conditions at te river mout. Results sow tat te φ TOTAL term was strongly dependent on te density gradient induced by river flow, especially during spring and early summer, wen te operation of Kerkini lake reduces stratification by approximately 14-59%. Te mean annual water column potential energy (φ TOTAL ) was also reduced by approximately 13%. stratification, water column dynamics, potential energy, Strymon River. 1. INTRODUCTION Stratification conditions of te water column in estuaries reflect te continuous competition between te stratifying influences of vertical gravitational circulation induced by buoyancy inputs and te mixing produced troug mecanical stirring by winds and tides. Furtermore, stratification appears to be a key environmental parameter for estuarine ecosystems, since it controls te vertical distribution of water properties (eat, salt, nutrients, pytoplankton and dissolved oxygen). In earlier papers (Simpson et al., 199, 1991; Nunes Vaz, 199; Nunes Vaz & Simpson, 1994; Lund- Hansen et al., 1996) te φ concept based on te potential energy of te water column, was used as a tool to evaluate te parameters, wic control te development and breakdown of stratification. However, most of tis researc was implemented in self sea areas dominated by strong tides, or in coastal areas wit low to moderate freswater flow. Tis paper attempts to quantify te mecanisms responsible for te vertical water density variations and te stratification mixing incidents at te mout of a dam controlled river (Strymon river) in a semi-enclosed bay (Strymonikos Gulf, Nortern Greece). Te model was run under two different scenarios (te influence of Kerkini dam and te absence of Kerkini dam) to assess te impact of water retention on te stratification conditions at te river mout.

2 G. Sylaios et al.. STUDY AREA Strymonikos Gulf is a semi-enclosed coastal water body in Nort Aegean Sea, aving a surface area of km and a water capacity of m 3 (Figure 1). Te area is considered among te most important nursery and fising grounds of Nort Aegean Sea for pelagic species (Kallianiotis, 1999). Strymonikos Gulf is te final recipient for te catcments of River Strymon (drainage area: 18,39 km ), wic outflows on te nortern part of te gulf wit a mean annual discarge of 59.5 m 3 s -1, supplying te gulf wit freswater and domestic, agricultural and industrial effluents (Koukouras et al., 1984; Vouvalidis, 1998; Parissis et al., 1; Sylaios et al., 1999; Sylaios, ; Haralambidou et al., 3, 5). Te discarge pattern of Strymon River sows strong seasonal variability, ranging on te average from 18 m 3 s -1 in August to 1 m 3 s -1 in April (data before Kerkini Lake; Mertzanis, 1994). River flow and water quality of Strymon is closely associated wit Kerkini Lake, a man-controlled artificial reservoir located 77 km upstream its mout (Tryfon et al., 1996). To cover te great demand in irrigation water from te nearby agricultural fields, te lake s dam remains closed from July to September, tus diminising river discarge from near zero to m 3 s -1. During recent decades a 3% reduction as been observed in te total freswater input to te Gulf due to reduced precipitation and extensive irrigation witin te drainage basin (Dounas and Koukouras, 199). Tis leads to te intrusion of a salt wedge, wic moves as far as 6-8 km upstream during low flow conditions (Parissis et al., 1; Haralambidou et al., 3). NEA KERDILIA OFRINION 4 48' '' 4 1' '' B ORFANI 5 km NEA BRASNA STRYMONIKOS GULF STAVROS OLYMPIAS STRATONION IERISSOS GULF IERISSOS NEA RODA 4 1' '' 3 38' '' Figure 1. Map of Strymonikos Gulf.

3 European Water 13/14 (6) 3 3. METHODOLOGY AND DATASETS Te φ concept is based on te potential energy of te water column, since te fully mixed water column contains more potential energy tan te stratified. Te amount of energy required to obtain a fully-mixed water column (φ, J m -3 ) is given by (Simpson & Bowers, 1981): 1 φ = ( ρ ρ) g z dz ; = 1 ρ ρ dz (1) were g is acceleration due to gravity (9.81 m s - ), ρ is water density (kg m -3 ), is water column dept (m) and z is te vertical coordinate (m). Mecanisms responsible for water column mixing are tidal stirring at te bottom and wind stirring at te surface. Tidal stirring on te bottom decreases φ by (Simpson & Hunter, 1974): dt T ε k = b ρ w u 3 b () were ε is an empirically determined coefficient of mixing (.38), k b is a bottom drag coefficient ( ), ρ w is water density (kg m -3 ) and u b is te bottom velocity (m s -1 ). Wind stirring at te surface decreases φ by: dt W δ k = S ρ W a 3 (3) were δ is a second empirically determined mixing coefficient (.39), k s is a surface drag coefficient ( ), ρ α is te air density (1. kg m -3 ) and W is te wind speed (m s -1 ). On te contrary, te stratifying mecanisms of te water column are te solar eating and te freswater buoyancy flux. Te solar eat flux at te sea surface increases φ by (Simpson & Bowers, 1981): dt α g Q = c p (4) were α and C p are termal expansion coefficients, wit α = o C -1 at 9 o C, C p = J kg -1 o C -1 and Q is solar eat flux at te sea surface (W m - ). Te freswater buoyancy flux, and tus canges in φ induced by estuarine circulation, is given by (Nunes Vaz & Simpson, 1994): dt R = 1 3 g N Z 4 ρ W ρ x (5) were is te water dept (m) and N Z is te vertical eddy viscosity coefficient (m s -1 ), varying wit tidal amplitude as N Z = γ u, wit γ = , and u te dept mean tidal speed (m s -1 ). It occurs from te above tat te stratification mixing model of te water column considers tat canges in time of te potential energy of te water column are due to wind mixing, eating, estuarine circulation and tidal currents, as follows:

4 4 G. Sylaios et al. 3 W = + α g Q 1 g δ k S ρα + dt W,, R, T CP 3 N Z 4 ρ ρ x ε k b 3 u ρw (6) In order to solve eq. (6) and derive te relative impact of eac individual term in te cange of te potential energy of te water column at te mout of Strymon River estuary, we used: a) meteorological data sets (daily mean wind speed and daily mean solar eat flux for te year, as provided by te NOAA ARL Real Time Environmental Application Internet site (ttp:// b) mean montly estimates of te ρ/ x-term, obtained by a series of CTD profiles at te mout of Strymon River, under different river flow conditions (Haralambidou et al., 4, Figure ), and c) daily mean values of tidal mean speed at te mout of Strymon River, as obtained by a two-dimensional tidal numerical model of Strymonikos Gulf, wic accounted for te fortnigt variability under neap and spring tidal conditions (Sylaios, ). To assess te impact of water retention from Kerkini lake, we used te mean montly values of Strymon river discarge during te period (Hatzigiannakis, 1999), as recorded before Kerkini and after Kerkini river dam. Different river flows vary te density gradient at te mout of Strymon River and influence te stratification conditions in te area. 6.E-3 5.E-3 Density Gradient (kg m -3 m -1 ) 4.E-3 3.E-3.E-3 1.E-3 Density Gradient =. (River Discarge) -.8 R =.53.E River Discarge (m 3 / sec) Figure. Density gradient variability, obtained from various CTD transects at te mout of Strymon River, in relation to river discarge. Line of best fit (linear approximation) was used to estimate density gradient values under different river flow conditions. Terefore, te time-interval witin wic te cange of te potential energy in te water column was examined was 1 day, and tus eq. (6) was integrated wit respect to time: t t t t α g 1 g ρ 1 3 φ TOTAL = δ ks ρa W dt + Q dt + dt ε kb ρw u dt (7) C 3 N ρ x p Z 4. RESULTS Results concerning te mecanisms responsible for water column mixing are sown in Figure 3. Wind circulation in te area appears influenced by moderate to strong norterly winds during te winter, and souterly winds in te summer. Mean montly water column potential energy (φ) due to wind influence was obtained daily from te first term of eq. (7). It occurs tat te wind term as an annual mean value of 7.7 J m -3, ranging between 1.4 J m -3 during October and.6 J m -3 during December (Figure 3b). Daily mean values of tidal current speed at te mout of Strymon River are presented in Figure 3c, as provided by te M tidal numerical model. Te mean montly values of

5 European Water 13/14 (6) 5 water column potential energy (φ) due to tidal influence were computed from te fourt term of eq. (7), sowing te almost constant mixing impact of tide (mean: 9.5 J m -3 ) on te water column dynamics (Figure 3d). Te variability of te mecanisms responsible for water column stratification is sown in Figure 4. Incident solar radiation varies seasonally in te area from 48.5 W m - in December to W m - in June (Figure 4a). Tis variability affects directly te water column stratification conditions at te mout of Strymon River by inducing a seasonal termocline. Te produced φ - eat term, as obtained from te second term of eq. (7), ranges between 8.9 J m -3 in December and 15.8 J m -3 in July (Figure 4b). Freswater river discarge induces a buoyancy input in te area, wic increases te orizontal density gradient and stratifies te water column. Te montly mean values of te density gradient at te mout of Strymon River, as obtained by following te equation produced in Figure, sow low values (. 1-4 ) during te summer monts (July to September), wen freswater input was limited due to te water retention by Kerkini dam. Te orizontal density gradient term appeared iger ( ) during Marc and May, wen te mean montly freswater discarge reaces a maximum (91. m 3 s -1 and 97.6 m 3 s -1 respectively). By considering river discarge as recorded before Kerkini Lake, we obtain iger density gradients ( ) in April, sligtly increased summer values ( ) and lower autumn and winter values ( ; Figure 4c). Water column potential energy (φ) sows a similar to density gradient beaviour under bot conditions. Under te influence of Kerkini Lake, low values (.1 J m -3 ) during July and August and ig values (7.9 J m -3 ) in May were computed. Te influence of water retention in stratification appears enanced during spring and early summer (April to June), wen water column potential energy is reduced by 59%, 14% and 15%, respectively (Figure 4d). In October te water retained in Kerkini dam is released, tus increasing te stratification of te coastal water column by approximately 5% (from 48.5 J m -3 witout Kerkini Lake to 73 J m -3 wit Kerkini dam). 3. (a) 5 (b).5 Wind Speed (m s -1 ) Φ - wind (J m -3 ) Tidal Current (m s -1 ) (c) Φ - Tide (J m -3 ) (d) Days Figure 3. Variability of mecanisms responsible for water column mixing. (a) Mean montly wind speed, (b) potential energy of water column (φ) due to wind influence, (c) daily mean tidal current speed, and (d) potential energy of water column (φ) due to tidal currents, at te mout of Strymon River.

6 6 G. Sylaios et al. Solar Radiation (W m - ) Density gradient (kg m -3 m -1 ) (a) (c) Φ - eat (J m -3 ) Φ - River (J m -3 ) (b) (d) Figure 4. Variability of mecanisms responsible for water column stratification. (a) Mean montly incident solar radiation, (b) potential energy of water column (φ) due to solar eat, (c) mean montly density gradient (open circles: witout Kerkini lake; solid squares: wit Kerkini lake), and (d) potential energy of water column (φ) due to river discarge (open circles: witout Kerkini lake; solid squares: wit Kerkini lake), at te mout of Strymon River. Table 1 presents te average, maximum and minimum values of te potential energy of te water column (φ) for eac term of equation (7) and te time of occurrence (mont). Table 1. Average, maximum and minimum values of te total potential energy of te water column (φ T ) and of te potential energy induced by solar insolation, wind, river discarge and tidal currents. Numbers in parentesis represent te monts of occurrence. φ TIDE (J m -3 ) φ WIND φ HEAT φ RIVER (J m -3 ) φ TOTAL (J m -3 ) (J m -3 ) (J m -3 ) wit Kerkini Lake Mean Minimum -5.7 () -1.4 (1) 8.9 (1).1 (7, 8) 4.8 (1) Maximum -3.3 (5) -.6 (1) 15.8 (7) 7.9 (5) (5) witout Kerkini Lake Mean Minimum 1. (8).9 (1) Maximum 11. (4) 1.7 (5) Figure 5 presents te temporal variability of (φ TOTAL ) term under bot scenarios, witin a typical year, as obtained by solving eq. (7). A gradual increase of water column potential energy, and terefore water column stratification conditions, from January to Marc, induced by te freswater input of Strymon River (mean discarge 67 m 3 s -1 ). Wen Kerkini lake is considered, mean montly river discarge reduces in April to 1 m 3 s -1 and te (φ TOTAL ) term decreases to te level of 73.9 J m -3. During te summer (May to August), enanced stratification conditions of te water column prevail (φ TOTAL from 98 to 18 J m -3 ), due to te influence of solar eat (φ HEAT from 13 to 15 J m -3 ). During te same period, te stratifying influence of river input reduces rapidly from 7.9 to.1 J m -3, since freswater discarge diminises to a minimum flow of 5 m 3 s -1 at te mout of Strymon River. From September to December te potential energy of te water column at te mout of Strymon River decreases gradually, since te mixing mecanisms of te wind and tide obtain maximum values. In December, te (φ TOTAL ) term obtains te minimum value (4.8 J m -3 ), due to increased mixing induced by te wind sear stress. Figure 5 also presents te stratification

7 European Water 13/14 (6) 7 conditions at te mout of Strymon River wen Kerkini lake is absent. Increased stratification conditions during spring and early summer period (April to June) are sown, producing a mean annual increase in te (φ TOTAL ) term of te order of 13%. 5. CONCLUSIONS A stratification mixing model, in terms of te potential energy of te water column (φ) as been developed and applied in Strymon River mout. Te model intents to assess te influence of water retention by a man-controlled artificial reservoir (Kerkini lake) on te stratification conditions at te mout of Strymon River. Te model accounts separately te effects of wind, solar eating, tidal circulation and river flow on te stratification mixing conditions in te self zone. Te model utilizes field data (CTD transects at te mout of Strymon River), istorical data of river discarge, wind and solar eat, and results of a tidal numerical model of te area. Results sow tat water retention by Kerkini lake plays an important role in te water column stratification conditions at te mout of Strymon River during spring and early summer period (April to June), by reducing te mean annual water column potential energy (φ TOTAL ) by approximately 13% (from 79. J m -3 to 91.5 J m -3 ). 5 Φ - Total (J m -3 ) Figure 5. Variability of te total potential energy of te water column (φ TOTAL ), trougout a typical year, at te mout of Strymon River (open circles: witout Kerkini lake; solid squares: wit Kerkini lake). REFERENCES Dounas C., Koukouras A.: 199, Circalittoral macrobentic assemblages of Strymonikos Gulf (Nort Aegean Sea). Mar Ecol; 13: Haralambidou K., Sylaios G., Tsirintzis V.A.: 3, Development of a numerical model to test alternatives to control saline wedge intrusion in an estuary. Proc. XXIII IAHR Congress; August 4-9; Tessaloniki, Greece, Teme A, pp Haralambidou, K., Tsirintzis, V.A., Sylaios, G.: 3, Testing alternatives for salt wedge management in an estuary wit te use of monitoring and a matematical model. Global Nest: An Intern. Journal; 5: Haralambidou K., Tsirintzis V.A., Sylaios G.K., Akratos C.: 4, Seasonal and spatial caracteristics of water quality in te estuary of Strymon River. Proc. of 7 t Int. Conf. on Protection and Restoration of te Environment, 8/6 1/7/4, Mykonos, Greece, CD-ROM, Section 3, No. 3. Haralambidou, K., Tsirintzis, V.A., Sylaios, G.K., Akratos, C.: 5, Seasonal and spatial caracteristics of water quality in te estuary of Strymon River. J. Marine Environm. Engin.;7(4): Hatzigiannakis S.: Hydrology of te Strymon River drainage basin, In Description of te Coastal Zone of Strymonikos and Ierissos Gulfs, Koutrakis E., Lazaridou E., (Eds.), NAGREF-GBWC, Kavala, pp. 5-3.

8 8 G. Sylaios et al. Kallianiotis A.: 1999, Te ancovy fisery in te Aegean Sea. A flourising industry or a lost affair?. Scient. Mar.; 6 (Suppl. ): Koukouras A., Voultsiadou-Koukouras E., Kattoulas M. : 1984, Bentic bionomy of te Nort Aegean Sea. I. Pysico-cemical caracteristics of te Strymonikos Gulf. Tal. Jucosl.; : Lund-Hansen L.C., Skyum P., Cristiansen C.: 1996, Modes of stratification in a semi-enclosed bay at te Nort sea-baltic sea transition. Est., Coast. & Self Sci.; 4: Mertzanis C.: 1994, Study of flooding waves at Strymon River, M.Sc. Tesis, Univ. of Tessaloniki, 88 p., (in Greek). Nunes Vaz R.A.: 199, Periodic stratification in coastal waters, In Modeling of Marine Systems, vol., A.M. Davies, ed., pp , CRC Press: Boca Raton, Fla. Nunes Vaz R.A., Simpson J.H. : 1994, Turbulence closure modeling of estuarine stratification. J. Geop. Res.; 99(C8): Parissis A., Sylaios G., Tsirintzis V.A. : 1, A numerical model for te study of salt intrusion at Strymon River mout, Nortern Greece. Proc. 1 st Int. Cong. on Ecol. Prot. of Planet Eart; June 5-8, Xanti, Greece, pp Simpson J.H., Hunter J.R.: 1974, Fronts in te Iris Sea. Nature; 5: Simpson J.H., Bowers D.G.: 1981, Models of stratification and frontal movement in self seas. Deep Sea Res.; 8: Simpson J.H., Brown J., Mattews J., Allen, G.: 199, Tidal straining, density currents, and stirring in te control of estuarine stratification. Estuaries; 13: Simpson J.H., Sarples J., Rippet, T.P.: 1991, A prescriptive model of stratification induced by freswater runoff. Est., Coast. & Self Sci.; 33: Sylaios G., Ioannidou D., Koutrakis E.: 1999, Water quality monitoring in Strymonikos gulf and gulf of Ierissos, N. Greece, In Water Pollution V, Brebbia, C.A., Anagnostopoulos, P., (Eds.), WIT Press, Soutampton, pp Sylaios G.:, A numerical model of coastal circulation and pollutants dispersion in Strymonikos Gulf. Proc. of 6 t Panell. Congr. in Ocean. & Fis., 3-6/5/, Cios, pp Sylaios G.:, A budget model of water, salt and non-conservative nutrients in Strymonikos and Ierissos Gulfs. Proc. of te 5 t Int. Conf. of EWRA; September 4 September 8, Atens, pp Tryfon E., Moustaka-Gouni M., Nikolaidis G.: 1996, Pytoplankton and nutrients in te River Strymon, Greece. Int. Revue ges. Hydrobiol ; 81 : Vouvalidis K.: 1998, Morpologic, sedimentary and oceanograpic processes and uman interventions contributing to te evolution of Strymon River estuarine system, P.D. Tesis, Univ. of Tesaloniki, 198 p. (in Greek).