Assign constituent models

You can assign constituent models by choosing Edit > Constituents > Assign Models.... The number of columns in the resulting window (Figure 139) will depend on the constituents you defined. Choose the desired model from the drop-down list in each constituent column.

Parameterise constituent models

You can parameterise constituent generation models by choosing Edit > Constituents > Parameters.... This can be done using parameter sets and is similar to that for rainfall runoff parameterisation. You can also use a Hazard Map, which requires the Spatial Data pre-processor plugin. This plugin is available through the Plugin Manager in the Tools menu. See Spatial data pre-processor for information on the plugin and Hazard Map Scaling.

Hazard maps are useful for informing catchment and land managers of parts of the landscape that are most vulnerable to certain environmental hazards, such as soil erosion or salinity. Scaling EMC and DWC values using a Hazard Map allows areas with "hazardous" land uses (eg highly grazed areas can be susceptible to higher levels of soil loss) to reflect the expected constituent magnitudes in such areas.

Constituent model types

The following types of constituent generation model are available in Source:

• Event Mean Concentration/Dry Weather Concentration (EMCDWC);
• Export rate;
• Nil constituent;
• Power function; and
• Power function with flow in millimetres.

EMC/DWC

The Event Mean Concentration (EMC) Dry Weather Concentration (DWC) model applies two fixed constituent concentrations (EMC & DWC) to a FU to calculate total constituent load in kg/unit time step. To apply this model you also need a rainfall runoff model with quick and slow flow proportions. The model is a scaling factor and is independent of the basic time step of the model. It is appropriate for estimating loads over longer time periods (at least monthly to annual scales).

The input data required include EMC and DWC values in mg/L (milligrams per litre).

Export rate

The Export Rate model applies a fixed constituent rate to a FU to calculate total constituent load in kg/unit time step. The model is not sensitive to changes in FU area. You do not need to incorporate a rainfall runoff model to apply this constituent generation model. It requires only a single parameter (load in kg per area per time step) and is useful for exploring sensitivity. The model is a scaling factor and independent of the basic time step of the model. It is therefore appropriate for estimating long term loads.

The input data is a single value for the export rate in T/ha/yr (tonnes per hectare per year).

Nil constituent

This does not apply any constituent generation model to a FU.

Power function

The Power Function is a model that fits a rating curve describing the relationship between constituent concentration or load and discharge in linear space. It generates constituent concentration in mg/L or load in kg/unit time step. Note that large errors in load estimates may occur if you extrapolate the relationship beyond the range of flows where measured data was available.

The power function requires a rainfall runoff model with total daily flow. Although these may be split into quick flow and slow flow components, they are not actually used by the power function model. Of the four constituent generation models described, the power function requires the most data but also provides the most sensitive constituent response to stream flow. Table 104 shows the input data format for the power function.

Power function (flow in mm)

The model is similar to the Power function model, but specifies discharge in mm rather than ML/day. See Table 104. This removes the dependency on area.

Configuring filter models

Filter models represent any transformation of constituents between generation within the FU and arrival at the link upstream of the sub-catchment node. Filter models process constituents within the FU and as with constituent generation models, are applied to FUs. Note that only one filter model can be applied to a sub-catchment/FU combination.

Just as with constituent generation models, filter models can be assigned and parameterised (optional) using the Geographic Wizard or the Edit menu commands.

Assign filter models

Choose Edit > Filtering Models > Assign Models... to assign filtering models to a scenario. Figure 141 shows the resulting window.

Parameterise filter models

There are two ways of assigning parameters sets to each filter model for any combination of sub-catchment, FU and constituent.

You should use the FU TEDI Preprocessor only when you specifically model the impact of farm dams within your catchment. For configuration details on the Farm Dams pre-processor, see Farm Dams. Use the next method when you have any other filter models applied.

The method for parameterising filter models using grid-based parameterisation is similar to that for rainfall runoff and constituent generation models.

Types of filter models

Seven types of filter models are available in Source, and they are described next.

1st order Kinetic Model k-C*

The 1st Order Kinetic Model k-C* filter model describes the decay or reduction in inflow concentration within a treatment facility such as a grass filter strip. It represents an exponential decay model, and can be used to simulate a range of processes, but has most frequently been used in evaluating the performance of constructed storm water treatment systems that have a surface storage, such as wetlands and ponds. It is equally valid in urban and non-urban catchments provided sufficient data has been collected to calibrate the decay rate (k) and the final background concentration (C*).

It is also the fundamental model used in music.

Farm Dams

This model works by capturing or filtering a proportion of runoff within each FU according to the total storage density of dams.

Overflow from dams in one FU will contribute to the total runoff of all FUs within a sub-catchment. The Farm Dam model is able to estimate the impact of farm dams on stream flow at catchment scales (up to several hundreds of square kilometres in area). There are several input data requirements that are required to set up a farm dam model. Refer to the Source Scientific Reference Guide fore more information.

Filter models act on constituents, so when you develop a scenario, you need to add a constituent that represents any runoff that is potentially captured (or could be captured) by farm dams (eg called "Flow"). The Farm Dams model will then be applied to this constituent. Alternatively, if you do not add a specific Farm Dams model constituent, you can still apply the Farm Dams model to any other constituent, regardless of name.

It is recommended that you do NOT model constituents in conjunction with modelling farm dams, as the Farm Dams model may adversely impact constituent loads and concentration calculations. If you wish to apply th model to an existing scenario, make a copy of the scenario before running the Farm Dams model, so that existing scenario water quality simulations are not affected by the altered runoff.

When modelling farm dams, you should initially use the Geographic wizard to create a scenario without the Farm Dams model. This scenario is then your "base case". You can then make a copy of the scenario, rename it, and apply the Farm Dams model to the copy of the original scenario and parameterise from the Edit menu. By having a separate base case and "farm dam" scenario, will allow the impact of farm dams on surface water runoff to be quantified.

Note There is no need to separately assign farm dam models to the "Flow" constituent in each FU. Farm dam models are applied to all FUs with a specified farm dam density when the pre-processor has been run. Therefore, you can skip the Assign Models step in the filter model.

To assign a farm dam model to a FU, choose Edit > Filtering Models > Assign Models.... In the constituent column, choose Farm Dam from the drop down menu. You can also decide which sub-catchments and FUs to assign this model to using the Map tab.

To parameterise the Farm Dams model:

• Choose Edit > Filtering Models > Parameters...;
• From the Available Methods drop-down menu, choose Farm Dam Pre-processor Figure 143;
• Select the flow constituent or other constituent that the farm dam model will be applied to;
• Select the functional unit that the farm dam model will be applied to. Each functional unit can have different farm dam parameters, such as different size class/volume relationships, demand factors, densities etc;
• Click on the Catchment Parameters tab;
• Set the total capacity of dams per unit area of FU, ie the number of megalitres per square kilometer.
• In the dam capacity-catchment area relationship table, specify the catchment area/volume relationship for each individual dam: for each dam, add one row to the table. This function represents the upstream catchment area corresponding to a 5, 10 and 100 ML dam (default). You can specify additional capacity groupings if necessary.
• To delete rows from the table, click the grey cell at the start of the row, then press Delete.
• To add rows to the table, click the grey cell containing the at the bottom of the table.
• Click on the Volume Parameters tab;
• Select the type of function to define the surface area/volume relationship;
• Specify the values for the surface area/volume relationship parameters A and B (Figure 144).
• Specify the farm dam size class/volume distribution function. This function will be used by the pre-processor to stochastically generate a sample of farm dams based on the density and size class distribution given in previous steps.
• The Dam Volume column can be changed to have any number of values by deleting or adding cells. To delete a cell, overwrite the current value with zero and press Enter.
• The Fraction column of the size class/volume distribution must sum to 1.
• Click on the Demand Parameters tab. Several constant values can be specified in Figure 145:
• Extraction threshold - this is the threshold below which no more water can be extracted from a farm dam (effectively the dead storage of a farm dam). The default value is 15%, which specifies that when the farm dam is at 15% of its total capacity, no more water can be extracted from the dam.
• Maximum typical volume of a stock & domestic dam. The capacity threshold between large irrigation dams and smaller stock & domestic dams (Default value is 5 ML).
• Demand factors - the proportion of water used as a proportion of dam volume. Some dams may be used frequently, and the water constantly replenished, whereas other dams may not be used at all. This proportion is the average usage factor for all catchment dams (Default value is 1).
• Monthly Demand patterns - extraction from farm dams are typically seasonal, thus water usage rates are modelled with a set of average monthly demand values. The monthly values must sum to 1.

If there are many FUs that require the Farm Dams model, set common parameters such as surface area - volume parameters, demand factors etc, for the first FU, then apply the parameters to all FUs using the Apply-to All FUs button. You can then customise the Farm Dams model for each FU, if necessary. Clicking Run will save all the parameters added to the preprocessor.

To remove the Farm Dams model for all FUs of a particular type, set the total capacity of farm dams per unit area of FU to zero, then click Run. The Farm Dam models will be removed from the currently-selected FU type.

Once all the values for the Farm Dam models for each FU have been entered, click Run to start the pre-processor.

If you change a parameter, and re-run the pre-processor, the previous model parameters are overwritten. If you wish to experiment with different parameters, make a copy of the "base" scenario, rename the copy, then change the parameters in the copied scenario.

Load based sediment delivery ratio

This model reduces the amount of sediment leaving a FU as a function of the amount of sediment generated in the FU. Essentially, the ability of a filter (such as a riparian zone) is limited, and if considerable load is applied to a filter, then trapping efficiency will drop to the point where all incoming material is passed straight through.

The model requires inputs of Sediment Loading Rate at Sill (SLRS), Sediment Loading Rate at Threshold (SLRT), stream length in metres and proportion of stream length affecting sediment delivery.

Load based nutrient delivery ratio

This model reduces the amount of nutrient leaving a FU as a function of the amount of load generated. Constituents in the slow flow remain unchanged. Although it is applied at the FU scale, this model can have catchment-wide effects.

Pass-through

This is the default setting for a filter model in Source. It preserves the amount of constituent generated within a FU and passes this amount to the sub-catchment node and implies that no filter has been applied.

Percentage removal

This model is a constant removal coefficient applied to the constituent load passing with baseflow (slow flow) and surface (quick) flow.

RPM Filter

The Riparian Particulate Model (RPM) quantifies the particulate trapping in riparian buffers through simulation of the processes of settling, infiltration and adhesion. It consists of 3 components that simulate particulate trapping:

• Coarse particulates by settling;
• Fine particulates by adhesion; and
• Fine particulates by infiltration.