Top 5 Flood Modeller Support Queries and How to Avoid Them
- 4 days ago
- 5 min read
Flood Modeller is advanced flood modelling software designed for accurate flood risk assessment, planning, and decision‑making, but even experienced users can run into technical challenges from time to time.
This blog highlights the top five most common Flood Modeller support queries and provides clear, practical guidance on how to resolve them, ensuring your models run smoothly. We've listed them in descending order of commonality, so keep reading to uncover number one.
5. How do I link the bottom of my 1D river network to my 1D urban network?
Solution: Link lines connect 1D river and urban networks. They are usually created using the 1D Urban-1D River linking tool. For more details, please refer to our "Linking 1D River Networks and 1D Urban Networks" help page.
The linking tool allows you to specify links between the active river and urban networks and generates a link‑reference (*.isl) file, which is a comma‑separated text file.
To link the bottom of a 1D river to a 1D urban network, you must first define a Head‑Time boundary with a unique label at the downstream end of the river network. When creating the link to the urban unit, select the linking type “Flow to Urban, Stage to River” in the linking table. After creating the link, use 'View defined linking' to confirm that it appears correctly on the map.
Once set up, run the model and check that flows behave as expected on both the river and urban sides. It may also be beneficial to carry out sensitivity testing on any link parameters.
For more complex configurations, please refer to the “Advanced Link‑reference Types” section of our help page.
4. How can I model an embankment breach in a 1D (or 1D-2D) model?
Solution: To model an embankment breach, connect a lateral SPILL unit to the RIVER section(s). A BREACH unit is then connected to the SPILL unit. When setting this up in the network table, ensure that the BREACH unit is inserted after the SPILL unit, not between the RIVER node and the SPILL unit.
As shown in the screenshot example below, the Upstream and Downstream labels in the BREACH unit must match those defined in the associated SPILL unit. All breach behaviour, including timing and geometry, is controlled through the parameters specified within the BREACH unit.
For further guidance and examples, please refer to our "modelling a breach" help page.
3. How do I reduce the Mass Balance Error in my 1D-2D linked model?
Solution: High mass balance errors in a 1D–2D linked model typically indicate an issue in the model build. This could be within the 1D model, the 2D model, or the linking between them. It’s important to note that any mass balance issues present in the standalone 1D model can be amplified once the model is linked to a 2D domain. As a result, all standard 1D mass balance troubleshooting steps remain relevant in a coupled 1D–2D setup.
Mass balance errors can arise for a variety of reasons, including convergence problems, interpolation issues, numerical instability, abrupt changes in channel geometry, inappropriate roughness values, or overly aggressive time‑stepping.
To understand which parameters may help reduce mass balance error in the 1D component, please refer to our knowledge base article.
You can also try enabling Double Precision (under Simulation → Backward Compatibility tab), which may help minimise errors caused by numerical rounding.
To spot the issue, you may investigate outputs such as:
FM1D link flow variable (default output)
High oscillations in the flow/water level (default output)
FM1D reservoir/river volume (enable via “Additional Output”)
Water level oscillations in the 2D domain
High Courant number / high Mass Error in the 2D Results
You may observe a significant difference between the Linked Error (m³) values reported for the 1D River and 2D in the model summary table. This indicates a discrepancy between the volume of water the 1D solver calculates as leaving the 1D network and the volume the 2D solver calculates as entering the 2D domain.
If the issue appears to originate at the 1D–2D link boundary, mass balance errors are often due to mismatches in bed elevation between the linked 1D nodes and 2D cells. Flow exchange between the solvers is based on water elevation. While the water surface elevation is consistent between 1D and 2D, differing ground elevations yield different water depths, leading to inconsistent flow calculations.
To investigate this, check whether the bed elevation from the 2D zcen check file at the linked cells (identified in the iwet check file) matches the bed elevation at the linked 1D sections. To extract elevation data at the left and right extremes of the 1D cross‑sections, you can use the Tabular CSV tool to export the left and right point coordinates and elevations to a .csv file for direct comparison.
2. How can I modify the roughness values on all the conduit units in my model? They include some irregular conduits.
Solution: While Flood Modeller includes a toolbox that supports many common batch-style tasks (such as “I want to edit all of my X nodes in Y way”), the Flood Modeller API offers a more flexible approach. By using Python, you can manipulate your model data programmatically and apply virtually any transformation you require.
To achieve this, we can break the task down into the following steps:
Open the .DAT file
Iterate over each unique conduit unit within the network.
a) For each conduit unit, multiply the roughness value(s) by a specified uplift factor
Save the updated network as a new .DAT file.
Below is a working example that demonstrates this approach. You can use it as a starting point and adapt it to suit your specific requirements.
1. I am getting a negative depth warning in my FM 2D model. How do I interpret or fix this?
Solution: As described on the help page, Error messages, this warning indicates that the 2D engine has detected cells with negative water depths during the initial mass-balance check. When this occurs, it requires further investigation at the specific locations and times where the issue is reported.
As a first step, consider increasing the negative depth threshold to 0.05 m or even 0.1 m. (0.1 m is the default value used in TUFLOW, whereas FM2D defaults to 0, meaning even very small negative depths, such as 0.0001 m, will be flagged.) This setting can be found at the bottom of the General tab in the 2D simulation window.
After setting the negative depth threshold to 5 cm, rerun the model and examine where the negative depths occur. Pay particular attention to whether they appear near or at boundaries, or within the middle of the floodplain. For example, in models that include H-links or elevation boundaries, negative depths near links or boundaries often indicate that the 1D model or boundary conditions are attempting to dry adjacent 2D areas.
Applying Proven Solutions
Many modelling issues stem from a small number of well‑understood causes, whether related to linking, stability, or parameter setup. By recognising these common problems and applying proven solutions, you can save time, reduce uncertainty, and improve overall model performance.
If issues persist, the Flood Modeller knowledge base and support teams remain invaluable for more complex troubleshooting.
.jpg)
