Rudi Rajar1, Matjaž Četina1, Fernando Gonzales–Farias2 and
Marina Pintar3
1University of Ljubljana, Faculty of Civil and Geodetic Engineering,
Hydraulics Department, Ljubljana, Slovenia
2Instituto Mexicano de Technologia del Agua, Cuernavaca, Mexico
3Water Management Institute, Ljubljana, Slovenia
Abstract: Many coastal lagoons in Mexico are used for marikulture (fish, shrimps, oysters). Regions near lagoons are usually exploited by agriculture. Drain water from the agriculture area often brings nutrients and pesticides into the lagoons, contaminating products of marikulture.
For a system of three coastal lagoons adjacent to the Carizo Valley (NW Mexico) a joint research project was initiated, with the main goal to propose measures to diminish contamination of the lagoons. First, several possible scenarios of improved agricultural management will be studied (different combination of type and amount of fertilizers and pesticides, timing of their application, method of application etc). For each scenario numerical simulations will be carried out by two models. Simulations by the GLEAMS model (Groundwater Loading Effects of Agricultural Management Systems) will give information on the inflow of contaminant in the drain water from the fields. Further on, simulations by two-dimensional hydrodynamic and pollutant transport model PCFLOW2D will show the transport and distribution of contaminant concentration in the lagoons.
By the means of both models and some measurements, used for the model
calibration, an optimum scenario of agricultural management will be proposed.
This should propose a solution, where the contaminant concentrations in the
lagoons will be below a permissible level, so that the products of marikulture
will be safe for consumption. The final scenario should be also be economically
acceptable.
This paper describes
the first applications of the 2D hydrodynamic model for the simulations of
circulation and water exchange in the lagoons and dispersion of the
contaminants, arriving from field drains. The lagrangian based particle tracking
method was used for this purpose.
Keywords: hydrodynamic modelling, 2-dimensional modelling, pollutant transport modelling, agricultural pollution, coastal lagoons, GLEAMS
There are numerous coastal lagoons in Mexico, where marikulture represents an important economical income. The main products are fish, shellfish and shrimps. The regions around these coastal lagoons are usually exploited by agriculture. In the regions of Sonora and Sinaloa (NW Mexico) there are more than 1.250.000 ha of agricultural area, which are irrigated and are situated in the neighborhood of 44 coastal lagoons with high fish production. The problem is that fertilizers and pesticides are used to increase the agricultural production, but the irrigation water from fields transports these contaminants into the lagoons, where they can contaminate marine food. The nutrients, resulting from different fertilizers, have damaging effect on the lagoon water quality, in greater concentrations they can cause eutrophication. Pesticides can often be toxic and their allowed concentration in coastal waters where marikulture is applied is usually very small.
One of the lagoon systems is situated by the Carizo valey in NW Mexico: the lagoons Bacorehuis, Jitzamuri, and Agiabampo (Figs. 1 and 2). The overall horizontal extent of the system is about 40 by 20 kilometers, but the mean water depth of Agiabampo lagoon is only 2 meters, and of Bacorehuis and Jitzamuri about 5 m. Only at the entrance channel from the Pacific the depth is near to 15 m. A preliminary study of modelling of the transport-dispersion of contaminants from agricultural regions has been carried out and is described further on. The main goal of the research is to diminish the concentration of nutrients and pesticides in the lagoons. This will be achieved by the following steps.
Fig.1 Scheme of the modelling methodology, linking of models GLEAMS and PCFLOW2D for the simulation of agricultural pollution in the lagoon system.
First, it is necessary to determine possible measures for diminishing the input of fertilizers and especially pesticides from the agricultural area to the lagoons. Several possible remediation measures will be studied, mainly in the form of improved agricultural management. The effect of these measures will be simulated by the numerical model GLEAMS (Groundwater Loading Effects of Agricultural Management Systems). Input data for the GLEAMS model are (Knisel et al., 1995): (1) Hydrologic data of the region (daily rainfall and temperature; monthly temperature, solar radiation, wind movement, dewpoint temperature), (2) Soil characteristics (3) Crop situation, (4) Pesticide characteristics and application data, and (5) fertilization and tillage data. The model output results are: (1) Water discharge (either surface or groundwater or both) into the lagoons, (2) Sediment outflow, (3) Nutrient outflow, and (4) Outflow of pesticides.
In the second step the transport and dispersion of nutrients and pesticides (referred further on as “contaminants”) inside the lagoon will be simulated by the means of a 2D numerical model PCFLOW2D. This will be done for each scenario of agricultural management and the spatial and temporal distribution of contaminant concentration in the lagoons will be determined, to find out if the concentrations are below permissible levels. In this way environmentally and economically optimum solutions will be determined.
This paper describes the second part of the research, i.e. the simulations of hydrodynamic circulation and contaminant dispersion in the lagoons.
At the University of Ljubljana, both two- and three-dimensional (2D and 3D) integrated models have been developed. For the case of Mexican lagoons, where depths are very small in comparison to the horizontal dimensions, and also the effect of thermal or salinity stratification is small, the 2D model PCFLOW2D is used.
Both 2D and 3D models have the following modules: (1) Hydrodynamic module, (2) Sediment transport module and (3) Transport-dispersion module. Bio-chemical modules for different contaminants are in development. Both models have already been used for solution of many practical application problems both in Slovenia and in other countries (Rajar et al., 1997). The 3D model has been used e.g. for the simulation of transport-dispersion of radioactive pollution in the Japan Sea (Četina et al., 2000) and for the simulation of transport of sediments and plutonium from the Mururoa lagoon (Rajar and Žagar, 1998) after the French nuclear tests in Polynesia.
The basic characteristics of the PCFLOW2D model are the following. The depth-averaged momentum and continuity equations are solved in Cartesian coordinate system by the finite volume method. A depth-averaged version of the well known k-e turbulence model (Rodi, 1980) is included. The hybrid scheme, a combination of central and upwind scheme, assures simplicity and robustness and it remains stable even with very complex geometry, and boundary conditions, although sometimes it involves a certain amount of numerical diffusion. For this reason a lagrangian based Particle Tracking Method which is free from numerical diffusion, is included in the transport-dispersion module.
Some preliminary simulations with the two-dimensional hydrodynamic and pollutant transport model PCFLOW2D have been carried out. The goal of this part of the study was to determine the circulation in the lagoons, exchange of water with the ocean, and simulations of the transport-dispersion of a contaminants, arriving with the drain water from the agricultural area.
Hydrodynamic circulation was simulated first. In the simulated Case G (Fig. 2) the following input data were taken into account. The main forcing factor is tide. Its amplitude is ±0.90 m and tidal period 12 hours approximately. West wind of 3 m/s, which was defined for summer conditions, was also taken into account, although the simulations have shown that its influence on the circulation is very small. Number of control volumes in the X and Y direction are 94 and 62 respectively. Space steps: Dx = 400 m, Dy = 333 m. The time step was Dt = 30 seconds.
The hydrodynamic circulation for this case is
presented in Fig. 2 by velocity vectors. At zero water level and at ebb tide the
maximum outflow velocities at the lagoon entrance are near to 1 m/s. In the
central Bacorehuis and in the Jitzamuri lagoon the direction of the flow is
reverse, which is due to the fact, that the previous tidal wave is still
propagating inwards. Some observations of the water levels in the lagoons have
roughly confirmed modelling results, but
different measurements of water levels and current velocities will be carried
out later during the research, for the calibration of the model.
Fig.2 Circulation velocities in the lagoons at ebb tide, at zero water level.
Fig.3 Simulation of contaminant dispersion from Drains 3 and 5 by Particle Tracking Method, 7 days after beginning of contaminat release. Continuos inflow of 8 particles per time step (=10 min).
Taking into account the hydrodynamic
velocity field from the described Case G, we further on simulated the
transport-dispersion of contaminants, which flow with the irrigation water into
the lagoons. As no actual data on their discharge and concentrations were yet
available, a continuous inflow of contaminants was simulated, represented by 8
particles per time step (this was taken 10 minutes in the transport-dispersion
simulations), which is flowing into the Bacorehuis lagoon via the surface drains
3 and 5. The results are presented in Fig. 3. The transport-dispersion module
simulated outflow of 8 particles per time step (10 min) from the drains 3 and 5.
The simulations can only show relative extent of mixing, because, as already
mentioned, the data on the realistic inflow of contaminants from the
agricultural area were not yet available. The value of the coefficient of
horizontal diffusion Dx = Dy was taken to be 0.1 m2/s. This was taken
from literature and experience, but as it strongly influences the results, this
value will be determined more accurately during further research, by
calibration, using measurements of tracer dispersion. But
the results on Fig. 3 show, that the inner part of the Bacorehuis lagoon is
badly flushed even after 7 days of tidal forcing.
In the second
simulated case, we wanted to determine how the lagoon water is exchanged with
the ocean during successive tidal cycles. We simulated an instantaneous release
of 20000 particles at the border lagoon-ocean, at the moment of zero water level
during ebb tide. Results are presented in Fig. 4. They indicate that after three
weeks the most contaminated parts of the lagoon system (Bacorehuis and Agiabampo)
are at least partially flushed. Lagoon Jitzamuri does not have almost any water
exchange with the ocean.
Fig.4
Simulation of penetration of the ocean water
into the lagoons and its dispersion. Simulation
of instantaneous release of 20.000 particles at the lagoon entrance at the time
of flood tide and zero water level.
Situation after 21 days.
The described hydrodynamic model has given some first information about the circulation and water exchange in the lagoons. To attain the final goal, i.e. to determine an optimum scenario of agricultural management, giving the least contaminant inflow into the lagoons, a combination of the GLEAMS and PCFLOW2D models will be used. For the calibration of the GLEAMS model, measurements of hydrologic, soil and other agricultural parameters will be necessary. The calibration of the PCFLOW2D model will demand measurements of hydrodynamic parameters (water levels, flow velocities) and dispersion parameters (measurements of tracer dispersion).
The first results show, that the described modelling methodology will be a very useful tool for solving the described problem of water quality in the lagoon system with marikulture activity.
Četina, M., Rajar, R., Povinec, P. (2000). Modelling of Circulation and Dispersion of Radioactive Pollutants in the Japan Sea. Accepted for publication in Oceanologica Acta.
Knisel, W. G., Leonard, R. A, Davis, F. M. (1995). Representing Management Practices in GLEAMS. Eur. J. Agronom. 4 (4), pp. 449-505.
Rajar, R., Četina, M., Širca, A.(1997): Hydrodynamic and Water Quality Modelling: Case Studies. Ecological Modelling, No. 101, pp. 209-228.
Rajar. R, Žagar, D. (1998): Transport of Sediments and Radionuclides from Mururoa Lagoon into the Ocean. Ch. 3.2 in: The Radiological Situation at the Atolls of Mururoa and Fangataufa, Technical Report, International Advisory Committee, pp 29-38.
Rodi, W. (1980): Turbulence Models and Their Applications in Hydraulics, IAHR Publication, Delft.