  |
Concept 4: freshwater lens in a coastal aquifer with a brackish lagoon
Concept 4. Freshwater lens in a coastal aquifer with a brackish lagoon
Consider an aquifer system of 5 km wide and 30 m deep. At the left side, the ocean is present, where the sandy aquifer contains saline groundwater. A sandy ridge with a width of 1 km exists, where during the wet season, fresh rainwater infiltrates. The natural groundwater recharge has been assumed to be some 520 mm (concentration is 45 mg Cl-/l) during half a year; during the dry season, no water infiltrates. A freshwater lens has evolved below the sandy ridge. Groundwater is abstracted for domestic purposes from many shallow wells. In landward direction we detect a lagoon of some 1.5 km width.
The model represents a typical situation with a wet and dry season. Local data have not been considered; they are unknown. Examples of unknown parameters are the exact groundwater abstraction rate, the natural groundwater recharge from rainfall, hydrogeological parameters like the hydraulic conductivity, the effective porosity, the salt concentration of the lagoon during both seasons, and the salt concentration of the infiltrated rainwater.
Figure 10: Conceptual numerical model of the coastal aquifer in Sri Lanka
Assumptions and model parameters:
- Hydraulic conductivity 10.0 m/day
- No anisotropy
- Effective porosity=0.35
- Lagoon waterlevel: wet season +1.5 m msl, dry season +0.5 m msl
- Lagoon salt concentration: wet season 300 mg Cl-/l, dry season 5000 mg Cl-/l
- Groundwater abstraction: 7 wells with 30 m3/year/m = 0.21 million m3/year/1000m
- Boundary conditions:
- Land side: hydrostatic pressure 4 m above lagoon, salt concentration top 45 mg Cl-/l, bottom 11,000 mg Cl-/l
- Sea side: hydrostatic pressure, seawater
- Bottom: no flow, depth -30 m msl
- Model parameters: 200 columns, 150 layers, time step 0.5 month
- Solute transport: 9 particles per cell initially implemented in each cell
- Longitudinal dispersivity 0.1 meter
- Transversal dispersivity 0.01 meter
Figure 10 shows the solute distribution in the coastal aquifer just before the tsunami. A freshwater lens of some 5 m thickness occurs in the sandy ridge. Upconing is present at the well locations where groundwater is abstracted (see yellow circles). Concentrations in these wells vary with the season: from some 100 mg Cl-/l at the end of the wet season to 450 mg Cl-/l at the end of the dry season (e.g. see well B. in figure 11 and 12).
The flooding during the tsunami has the following characteristics: duration 2 hours, head 3 m. Abstraction of groundwater is assumed to stop during one year. Figure 11 shows that the concentrations in the observation points A., C. and D. go up due to the infiltration of sea water during the tsunami. Well B. shows a decrease in chloride concentration in the first year after the tsunami because abstractions have stopped: the upconing of brackish groundwater 'reduces' due to density differences. After one year, the concentration goes up due to the start of the abstraction.
The intrusion shows as a brackish front moving through the freshwater lens (animation). The lens itself remains intact and does not change in size in the period without abstraction after the tsunami. Hydrodynamic dispersion causes the mixing of fresh, brackish and saline groundwater. Figure 12 shows the results of the simulation 0.5 year after the tsunami. The intrusion of the brackish water can be seen just below ground surface.
Figure 11: Chloride concentration as a function of time at various depths and at various positions in the freshwater lens. Flooding with seawater occurs at year 0. Seasonal variation in concentration is mainly caused by seasonal variation in recharge.
Figure 12: Position of the observation points in the groundwater system.
Conclusion
The tsunami affects the chloride concentration in the coastal aquifer, but increases in concentration are not spectacular. Within some years, depending on the rainfall, the increase in concentration is probably vanished. However possibly the situation modelled is not the worst case situation. When sea water remains in pools and/or local depressions, salt water intrusion will last longer. Less rainfall in the months after the tsunami will prolong the effect of the intrusion. Continued leaching of salts from the soil by infiltrating rainfall is not modelled. The effect of the mixing of saline water with freshwater due to the pressure of the wave is not considered.
Remark
Local circumstances just around the drinking water wells can not be modelled with this regional conceptual model; finer models with more detail should be used to simulate the effect of the tsunami on this scale.