Background and methodology

Results

Conclusions and further work

The objectives for this investigation carried out by Cardiff University School of Engineering (Westwater, Falconer and Lin 1999) were to produce an operational two dimensional model of the Fleet lagoon, calibrated against field data, and to establish the flushing characteristics of the lagoon using simulations of the transport of a conservative tracer. The model and tracer simulations could then be used to predict distribution of water quality indicators such as salinity and nitrogen distribution in the Fleet and from stream inflows.

The Fleet is characterised by unusual hydrodynamic properties, in that the velocities and depths at the head of the lagoon are very small which leads to low Reynolds number flows. This hydrodynamic phenomenon means that the effective bottom friction increases with decreasing Reynolds number (or reduced velocity x depth), thereby leading to increased headlosses. This effect is not currently included in numerical hydraulic models and was an important component of the current study. More extensive research is required in the future into studying the effects of this complex hydrodynamic phenomenon, particularly as it relates significantly to the transport of fluid mass and solute contaminant and water quality indicator fluxes in tidal wetlands.

The numerical model used for this part of the work is much more refined than that used for the initial budget modelling used by the EA. However, as with any model, there are still uncertainties included within the model which should be appreciated, including effects of turbulence and lagoon bed roughness on water flows in the lagoon, erosion and deposition of sediments, and chemical and biological processes relating to water quality parameters. The particular case of the Fleet, being very shallow with parts subject to tidal drying, and with highly variable width along its length, presents particular difficulties in developing a realistic hydrodynamic model. The model used did, however, provide reasonable agreement when calibrated against field data.

Bathymetric data obtained in 1998 for English Nature were used in the model, as the data originally used (from Robinson 1983) were found not to be sufficiently detailed. Data on bathymetry of the western part of the Fleet were, however, sparse, which compromises the accuracy of the model predictions somewhat. Tidal elevation data from Portland Harbour were initially used in the model, but were found not to give a good fit with the measured data on calibration of the model. Therefore the tidal elevation data from the 1960=s reported in Robinson et al (1983) were used (Westwater et al 1999).

Difficulties in fitting the model to the measured data were also encountered due to the relatively great effects of flow resistance of the lagoon bed in such shallow waters as occur throughout the Fleet. This difficulty was overcome by adjusting the values for bed resistance in the model to better match the measured data.

Flow velocities in the Fleet were found to be generally low (0.2 B 0.3 m/s) and one dimensional (i.e. along the axis of the lagoon), with the exception of around Smallmouth and the Narrows, where relatively high flows of around 1 m/s were observed. At Smallmouth just inside the Fleet there was a large eddy, which was apparent during both neap and spring tides. In the Littlesea area of the eastern Fleet there was also a more complicated flow regime, with transverse flows (i.e. across the Fleet) in all plots at low tide, corresponding to the tidal drying of large areas in this embayment with flows following deeper channels in the lagoon bed. At high tides the flow reverted to the predominantly one-dimensional flow along the Fleet as seen for other areas.

Tracer studies indicated that the eastern Fleet has good tidal exchange with Portland Harbour up to Chickerell Hive in the Littlesea area, particularly during spring tides. During neap tides, tidal exchange with the west Fleet was extremely weak, with very little tidal exchange from Littlesea westwards. Tracer studies simulating release of a tracer at Abbotsbury showed that over ten tidal cycles the tracer only travelled as far as the narrow section adjacent to Abbotsbury, and did not even reach Rodden Hive Point. Tracer studies simulating release of contaminants in streams also demonstrated that Abbotsbury is a potential problem area, as it has three streams (Coward=s Lake, Mill Stream (Abbey Barn) and Mill stream (Horsepool) (also known as Portesham Mill Stream)) flowing into it, but with very little tidal exchange. Again, contaminants did not travel further than the narrow section adjacent to Abbotsbury over ten tidal cycles. The other streams appeared to have little effect on the gross water quality characteristics as their flows are lower, and they enter the Fleet further eastwards where tidal exchange is better. Some localised effects were seen in the Littlesea area.

Hydrodynamically the Fleet may be considered as two separate sections:

East Fleet, from Smallmouth through The Narrows and into Littlesea, where there is reasonable tidal exchange and good tidal mixing due to the narrowing of the channel. In this area the stream inputs appear to have little effect on gross water quality characteristics, with localised effects at low tides around the point where East Fleet stream enters the lagoon.

West Fleet, from Littlesea westwards. This area has a significant lack of tidal flushing and circulation, with large areas of stagnant water, particularly at the Abbotsbury end of the Fleet. It appears that the western end of Littlesea, with its large areas drying out at low tide, acts as a barrier to tidal flushing from the east Fleet, resulting in low flows and lack of flushing, leading to high sensitivity to contaminant release, in the western section of the lagoon.

Recommended further work includes:

• simulations of salinity distributions and other water quality indicators such as dissolved oxygen, nitrates and phosphates;
• refinement of the model by further investigation of the influence of bed roughness, effects of lagoon bed vegetation and effects of tidal flooding and drying on the hydrodynamics of very shallow areas such as the Fleet;
• further refinement of the model by obtaining additional bathymetric data for the Abbotsbury area, including location of any channels which would affect flows, and to refine the grid area from 50m down to 33m or 25m, to better investigate two dimensional flows.

Next Section                 References