2D Simulations - Outputs Sub-tab

Upon running your simulation, a number of output files are generated; some are by default whereas others can be requested by the modeller. Outputs are detailing in the 'Outputs' sub-tab of the 2D interface.

A 2D model LOG file will always be created. By default, this takes the same name and is written to the same location as your model XML file. However, you can specify a different name and location on the General Tab of the 2D interface.

You can specify different outputs for each domain in your model (and these can also be different formats for each output). Each defined output in each domain will be written to a separate folder. Each new output folder is created in the location of your XML file when you run a simulation. These model outputs are defined in the 'Domains Output' table, located on the 'Outputs' sub-tab for each domain.

For each line in the 'Domains Output' table represents a separate defined output, i.e. a new folder will be created for each. Folder names take the name of your XML file followed by an underscore and a sequential number, e.g. my-project_01, my-project_02. Note that the sequential numbers 01, 02, etc. will match the row number in the table.

SMS Grid Format

If your chosen ‘Output Format’ is SMS Grids, the folder will contain a 2DM file and a SUP file with the ‘Output Name’ specified; these contain the model grid data and metadata (summary of variables output) respectively. It will also contain a number of DAT files, one for each of the selected 'Output Parameters'.

Filenames for these DAT files will take the specified ‘Output Name’ followed by an underscore and the name of the parameter, e.g. Output-domain 1_depth, and contain the relevant time series values.

To create a new set of SMS grid format outputs for a domain, highlight the domain of interest and click the ‘add’ button to bring up the following window:

Enter an ‘Output Name’ and ‘Frequency’ (in seconds) by clicking in the relevant field and entering free text. The ‘Frequency’ for each output specifies the times at which the selected variable(s) are calculated (which can be different from the model timestep). Ensure SMS Grids is selected in the ‘Output Format’ field and click ‘OK’ to add the output as a row to the 'Domains Output' table. Select this row in the table to highlight it, and then choose the parameters of interest by selecting the checkboxes in the 'Output Parameters' list:

XMDF Grid Format

The procedure for specifying outputs in XMDF format is the same as that for SMS grids. To specify this format, select the appropriate grid type from the 'Output Format' dropdown list:

Maximum Extent Grids

By default, new 2D model derivations will include the output of maximum extent grids in ASCII raster grid format. These will be generated for each of the default parameters; depth, water level, flow and velocity.

To add a max extent output to an existing model, first click the 'Add' button below the 'Domain Outputs' table. In the 'Add Output' pop-up window, select the format of your maximum extent data; these can be ASCII raster grid format or GeoTIFF grid format. Note that the GeoTIFF format will maintain any grid rotation present in your model domain, whereas for the ASCII format a conversion to “X-Y” orientation will be performed.

When one of these options is selected, the frequency value no longer applies and hence is disabled. Thus, just provide a name for your max extent output and click 'OK' to add it to the list of outputs.

You can then specify which parameters to create max extents for by ticking them in the 'Output Parameters' list.

CSV time series Format

If your chosen ‘Output Format’ is CSV time series, the folder containing the model outputs will contain a single CSV file. This will consist of a column of all the times at which calculations are made (as specified in ‘Frequency’) and then columns of the calculated values of each selected variable(s) at specified points. To create a new set of CSV time series outputs for a domain, highlight the domain of interest and click the ‘add’ button. In the pop-up window, enter an ‘Output Name’ and ‘Frequency’ (in seconds) by clicking in the relevant field and entering free text. The ‘Frequency’ for each output specifies the times at which the selected variable(s) are calculated. Adjust the ‘Output Format’ to CSV time series so the ‘Point Output locations’ section becomes enabled:

Points at which you desire the calculations to be made can be entered directly into the table. Alternatively, click ‘Add points from file’ to browse to a shapefile using standard windows explorer browser. Click ‘OK’ to add the output as a row to the 'Domains Output' table. Select this row in the table to highlight it, and then choose the parameters of interest by selecting the checkboxes in the 'Output Parameters' list:

For either 'Output Formats' chosen, rows in the 'Domains Output' table can be edited by double clicking on the row or clicking on the ‘Edit/View’ button below the table. The ‘Remove' button can be used to remove rows from the table; a pop-up box will appear asking for confirmation prior to deletion.

You can also specify additional custom outputs for a simulation. These can be mass balance files (text format), check files (raster grids, .asc), flood hazard outputs and flow totals crossing a specified line. It should be noted that Flood Modeller post-processing tools can also generate flood hazard and flow totals crossing a line data from your standard flow, velocity and depth outputs. More information about the setup of Mass Balance Files and how to use these can be found on the page Mass balance files and calculations.

Shear Stress Calculations

The key parameters related to shear stress are:

• bed shear stress
• excess shear stress
• shields parameter
• stream competence

Bed shear stress

Bed shear stress is the tractive force exerted on a surface (i.e.the river channel) by the flow of water. It has been proposed that the Bed Shear Stress equation (Thompson and Crooke, 2013) is used to calculate this variable for each cell.

The Bed Shear Stress equation is solved for each time step. This allows the results to be displayed as a time-series graph and a map which can be animated, and colour coded based on categorical values. The user can create point plots for any cells within the study area to provide a graph of results over time.

To solve this equation for 2D simulations the required dynamic inputs are:

• depth averaged flow velocity of each cell for each time step
• average flow depth in each cell for each time step

The constants that are required are:

• density of water - 1000 kg/m/s²
• acceleration due to gravity - 9.81 m/s²
• Manning’s n – this can either act as a global constant or a localised constant via a shapefile

The Bed Shear Stress equation does not account for the direction of force and hence shear stress is calculated as a scalar value as opposed to a vector. Furthermore, the formula has two important limitations that need to be communicated to the user. The first is that because the Bed Shear Stress equation uses flow depth as a divisor, an asymptote is created whereby the value of predicted shear stress increases rapidly as the flow depth tends towards zero. Typically, this is dealt with by setting a cut-off value of 0.1-0.12m below which shear stress is linearly reduced to zero (TUFLOW, 2018; Pasternack et al., 2006). Secondly, the Bed Shear Stress equation may not accurately represent shear stress for reaches with very coarse material such as cobble or boulders, although there is no clearly defined upper limit (USGS, 2008). Fine flocculent sediment can also cause deviations between measured and predicted values. Although this is still a less well understood area of sediment transport, modification factors to account for the different behaviour of fine sediment are typically not required if the grain size is greater than approximately 0.016mm (Stevens, 1991; Zou et al., 2017).

Excess Bed Shear Stress

Excess bed shear stress is used indicate whether the flow has enough tractive force to erode sediment. If excess bed shear stress is equal to or less than 0, then sediment is unlikely to be entrained. Conversely if excess bed shear stress is greater than 0, it is possible for sediment transport to occur. Excess bed shear stress is calculated as bed shear stress minus critical shear stress (Equation 1.2). Excess bed shear stress is computed for each time step. For the outputs, the value for each time-step is displayed as a time-series graph and a map that can be animated and colour coded based on categorical values. The user can plot the results at each point within the study area (creating point plot graphs). Because excess bed shear stress is derived directly from shear stress, the same limitations apply, mainly that results become unreliable for flow depths less than 0.1m and for particularly fine or coarse sediment.

Equation 1.2 Excess Shear Stress – Habibi, 1994

Where

To solve equation 1.2 critical shear stress first needs to be calculated. Critical shear stress represents the forces preventing sediment from being entrained by the flow. Typical solutions range from 0.0378-223 depending on the sediment grain size (USGS, 2008).

Equation 1.3 Critical Shear Stress – USGS, 2008

Where

Notes about the critical shear stress calculation:

• For solving the Bed Shear Stress equation, Water density and sediment density are both constants, with sediment density typically assigned a standard value of 2650 kg/m3.
• For Dr (the representative sediment size), the user is able to specify a single value for the whole study area, or upload a 2D shapefile that contains different values for Dr that are spatially georeferenced over the study site. For the case in which a single value of Dr is specified, critical shear stress will act as a constant over the whole study area. For the case in which Dr varies spatially, tcr will act as a localised constant, but in both instances the value does not change with flow conditions. In most instance Dr will be used to represent the average grain size (denoted as D50). However, the more generalised case of Dr has been used here because it will allow the user to specify any grain size, allowing for the effects of the flow on any sediment grain size of interest to investigated.
• Dimensionless critical shear stress (t*c) is a constant that needs to be specified by the user – with a value normally between 0.03 and 0.06 and default value of 0.047 (Habibi, 1994).

Critical Shear Stress and the Shields Parameter (see below) are valid for large grain sizes, but the values of these variables approach a minimum of around 0.054 for coarse cobble (Figure 1.1) (USGS, 2008; Bottacin-Busolin et al., 2008). For fine sediment, results start to become less accurate at approximately 0.016mm as sediment start to exhibit cohesive behavior (Zou et al., 2017).

Figure 1.1 Critical shear stress by particle size classification for determining approximate condition for sediment mobility

Shields Parameter

The Shields parameter (also referred to as the shields number of shields stress), is a non-dimensional number used to predict critical shear stress for particle entrainment for a given clast of sediment (Equation 1.4). Computed values typically range from 0.025-0.8 depending on the grain size of the sediment (USGS, 2008).

where

The Shields parameter is calculated from the density of sediment and water (both of which are constants), acceleration due to gravity (a constant), representative grain size which will need to be input by the user, either as a constant or as a shapefile and bed shear stress. Because bed shear stress is used in the calculation of the Shields parameter, the value will vary with the stage of flow.

Stream competence

The stream competence is the maximum particle size that can be carried by a river or stream (Equation 1.5).

Equation 1.5 Stream competence - Vázquez-Tarrío et al., 2019

where

Similar to the Shields parameters, bed shear stress is used to calculate the entrained grain size. As such, the value will vary with the stage of flow. The outputs for stream competence are time-series graphs, 2D maps that can be animated and point plots. The maps are colour coded based on categorical values.

References

Bottacin-Busolin, Andrea, Tait, Simon, Marion, Andrea, Chegini, Amir, and Tregnaghi, Matteo, 2008, Probabilistic description of grain resistance from simultaneous flow field and grain motion measurements: Water Resources Research, v. 44, no. 9, WO9419, 12 p.

Habibi, M., 1994, Sediment Transport Estimation Methods for River Systems, University of Wollongong

Stevens, R.L, 1991, Grain-size distribution of quartz and feldspar extracts and implications for flocculation processes, Geo-Mar. Lett. 11 (3) (1991) 162–165.

USGS, 2008, Simulation of Flow, Sediment Transport, and Sediment Mobility of the Lower Coeurd’Alene River, Idaho, United States Geological Survey.

Hickin, 2010, River Geomorphology, Colby University

Vázquez-Tarrío, D., Fernández-Iglesias., E., García, F.M., Marquínez, J., 2019, Quantifying the Variability in Flow Competence and Streambed Mobility with Water Discharge in a Gravel-Bed Channel: River Esva, NW Spain, Water, 11, 2662

Zuo, Liqin, Roelvink, Dano J.A., Lu, Yongjun, Li, Shouqian. (2017). On incipient motion of silt-sand under combined action of waves and currents. Applied Ocean Research. 69. 116-125.