next up previous contents
Next: Additional display options Up: Analysing model results Previous: Analysing model results

Subsections

PEC Calculations

As GREAT-ER introduces a new method to estimate environmental concentrations (calculations are made on real-world data instead of large-scale generic data sets) with frequency distributions, the calculation of PEC values must be different from the known methods.

The PEC values calculated by GREAT-ER are currently under discussion. For this reason not just one but a collection of values are provided which are divided into two items: PEC initial and PEC catchment. The underlying theory and discussion would exceed the limitations of this manual. We refer readers to the article in Chemosphere (May 1999: in press).

PEC initial

is the spatial aggregation of concentrations in the river directly after emission. This is comparable to PEClocal as defined in the EU Technical Guidance Documents, although its calculation is different. Based on the C sim, start mean values, the mean is calculated for all stretches receiving treated or untreated discharges:

\begin{displaymath}PEC_{initial} =
\frac{\sum_{i\in \cal{R}} C_{sim, start, i}}
{n}
\end{displaymath}

with

\begin{displaymath}{\cal R} = \left\{ C_{sim, start, 1}, \ldots, C_{sim, start, n} \right\}
\end{displaymath}

C sim, start, i is the concentration at the beginning of river stretch i, directly receiving treated or untreated waste water emissions.

By definition, the PEC initial is calculated on the base of the C sim, start mean values. Selecting the item PECinitial the calculation starts.

Please note that PEC calculations are not available if a subcatchment is selected.

After the calculation a dialog displays the result alongside additional information.

PEC initial Calculation Results

Figure 2.24: PEC initial calculation results  

PEC catchment

is the aggregation of concentrations representative for the entire catchment. Although this is a new concept typical for geo-referenced exposure assessment, it could be compared to the PEC regional in the EU Technical Guidance Documents.

As it is an average value, C sim, internal is the "most representative" value for the concentration in the stretch. As PEC catchment aims to provide a measure for the "representative" concentration over the whole of a catchment, it is based on the most representative local values. As mentioned above, the PEC catchment is currently under scientific discussion. To demonstrate the variability of results based on the various definitions, GREAT-ER computes four different PEC catchment values. These values differ in the set of stretches selected for the calculation and in the weighting of concentrations.

All PEC catchment values are defined as the mean of the weighted concentrations of the selected stretches. Possible selections are (A) all stretches within a catchment or (B) only polluted stretches. Option (A) is more comparable to the current regional exposure assessment approach (unit world models) in which all surface waters in a region are considered for the dilution of the chemical mass loading. Clearly different, option (B) considers only the loaded river stretches. The intention of this approach in the meaning of risk assessment is to focus on the river stretches potentially at risk: the loaded stretches.

Both approaches have to deal with different issues concerning scale dependencies or data requirements. These items are discussed in the referred article.

1.
Weighting by stretch volume for both stretch selections. The volume is calculated (assuming a rectangular cross section) by flow, flow velocity and length. The weighting method focuses on large rivers and therefore on high dilution situations where exposure (and risk) levels may be lower than the representative rate for the entire catchment.

2.
Weighting by stretch length for both stretch selections. The interpretation of the weighting method is that stretches with equal length are of equal importance. Small rivers are considered to be equally valuable as large rivers.

Methods 1 and 2 depend on the scale and detail of the underlying digital river network, and the more (unpolluted) headwaters are included, the lower the aggregated PEC catchment will be. Conversely, in a low detailed river network, small (and possibly more valuable) stretches are neglected.

3.
Weighting by flow increment considers all stretches within a catchment. The weighting of a stretch is calculated by the difference of flow in relation to its upstream stretch (two stretches in the case of a confluence). Thus the values related to stretches are similar to those in weighting by length. A slight accentuation might be given to receiving stretches because, besides the natural flow increment, these are artificially influenced by the waste water flow.

The initial stretch of a river is considerd with its complete flow (there is no longer an upstream stretch and thus the increment is the flow). If these stretches are unpolluted, the weighting method is more or less independent of the scale of detail of the digital river network: no matter how detailed rivers are digitised, the flow of a given stretch can be considered as the sum of all its upstream stretches and therefore the stretch representing the headwaters is of the same value as all stretches regarded separately.

The calculation of PEC catchment can be based on any of the previously derived percentiles of the model results. When selecting PECcatchment the existence of previously calculated values is first checked because, depending on the catchment extent and detail of the digitised river network, the calculation will take several minutes.

PEC catchment Selection

Figure 2.25: PEC catchment selection  

A progress bar will index the progress of the calculation. After completion the values mentioned above are listed in a small report window:

PEC catchment Calculation Results

Figure 2.26: PEC catchment calculation results  

For additional analysis tools see the section below und Chapter 3 (Menu Structure).


next up previous contents
Next: Additional display options Up: Analysing model results Previous: Analysing model results
F. Koormann & J.-O. Wagner, Institute of Environmental Systems Research, University of Osnabrück