Nutrient load analysis and modelling

Background and Methodology


Background and Methodology

Nutrient load analysis and simple modelling of the Fleet lagoon was carried out by Environment Agency using data obtained during water quality monitoring studies from 1996 to >98 (Murdoch 1999) and information from Mainstone and Parr 1999. The relative contributions of point sources (sewage works discharges), freshwater and wildfowl contributions to the nitrogen and phosphate concentration levels in the lagoon was quantitatively assessed. The modelling attempted to assess input loads on daily and seasonal as well as annual timescales.

Before the modelling was carried out, nutrient sources were identified and their loads estimated. Two direct discharges from Wessex Water plc. sewage treatment works (STW) at Abbotsbury and Langton Herring discharge into streams entering the Fleet (Mill stream (Horsepool) and Rodden Stream respectively). Other smaller sewage effluent discharges were not included in the model as they are very small, and their input is not thought to be significant in overall terms (Mainstone & Parr 1999).

No continuous flow monitoring data were available for the streams, therefore flows were estimated from mean flow statistics using the Institute of Hydrology=s Micro Low Flow methodology. Estimated daily stream flows are then calculated using flow data from the nearby Broadwey gauging station to estimate daily variations. Calculated flows were compared to spot gaugings of stream flows, with relatively good agreement between the two for all streams except Mill Stream (Horsepool) and Mill Stream (Abbey Barn). Flows for Mill Stream (Horsepool) were adjusted in the light of this comparison as it receives effluent from Abbotsbury STW which would not have been taken account of in the flow estimations. Nutrient monitoring data were available for six of the seven streams (inputs from Herbury stream were assumed to be similar to those of the nearby East Fleet stream).

Diffuse input flows (run-off from land into the lagoon directly) were estimated using Micro Low Flow methodology, and were assumed to be evenly distributed between Abbotsbury and East Fleet. Nitrogen concentration of diffuse sources was assumed to be 5 mg/l N, estimated from the concentration of nitrogen found in borehole water nearby. No data were available for phosphorus content of borehole water, and since phosphorus mobility is low, diffuse inputs of phosphorus were ignored for the purpose of this modelling.

Estimated inputs from faecal matter of wildfowl, including swans, were put into the model using estimates based on monthly numbers of birds and the nitrogen and phosphorus content of faeces (Mainstone & Parr 1999). For the purposes of the modelling, the worst case of all wildfowl faeces input to the lagoon was assumed (rather than one third defaecated into the lagoon with two thirds onto land as estimated in Mainstone & Parr 1999).

The model (ECoS (2)) was used to obtain simulations of annual budgets. Use of this model means that short term accuracy is traded for the ability to do long simulations. Physical and tidal data required by the model were taken from Robinson=s work (Robinson 1983; Robinson et al 1983). This model is much more approximate than that developed by Westwater, Falconer and Lin (1999), because of the level of accuracy of the bathymetry, tidal and flow data used, and the fact that it is a one dimensional model rather than two dimensional.



For total inorganic nitrogen (TIN), the freshwater streams were found to dominate inputs to the system. Stream inputs showed a strong annual cycle with winter highs and summer lows. The inference is that winter runoff and flushing of nitrates is responsible for the winter peaks observed. Annual loads for three of the streams (Abbey Barn, Rodden and West Fleet) were around 20-25 tonnes N per year, with the other streams (Horsepool, Herbury and East Fleet) around 15 tonnes N per year and Cowards Lake lowest at 7 tonnes N per year. The loads for Mill stream (Abbey Barn) and for Rodden stream include the inputs from Abbotsbury and Langton STWs respectively, although both are shown separately in addition on the graph below. Abbotsbury sewage treatment works (STW) output and wildfowl had very low loadings (2 tonnes N per year). Input from Langton STW was negligible. Total inorganic nitrogen (TIN) loads were greatest at the Abbotsbury (western) end of the Fleet.

Annual nitrogen loads to the Fleet (from Murdoch 1999)


The model showed a double annual cycle in phosphorus loads, with winter highs and summer highs. The winter highs appeared related to the stream inputs (particularly Mill stream (Horsepool) and Mill stream (Abbey Barn), but also Cowards Lake, Rodden stream and West Fleet stream). The stream loads from Mill stream (Horsepool) and Mill stream (Abbey Barn) are highest at about 8 kg/day in winter. Mill stream (Horsepool) load includes the effluent from Abbotsbury STW (which contributes approximately 2 kg/day). Other stream inputs were much lower with winter inputs of less than 2 kg/day. The summer highs appear to be related to the wildfowl inputs; the impact of wildfowl may be exacerbated owing to associated high organic loadings resulting in increased anoxia within sediments and enhanced phosphate release (Mainstone pers. comm.). However, there could be effects from Abbotsbury STW inputs entering the western Fleet and not being flushed out to the eastern Fleet and the sea.

Annual phosphorus loads to the Fleet (from Murdoch 1999)

Next Section                 References