Difference between revisions of "TUFLOW Remapping"

From Tuflow
Jump to navigation Jump to search
Line 1: Line 1:
 
= Introduction =
 
= Introduction =
With the release of TUFLOW 2020, the combination of Quadtree mesh and Sub-grid Sampling (SGS) method has offered a great flexibility for flood model building. Now a wide range of mesh size can be applied in a model based on the geographic feature and the area of interest in a project. Especially with SGS method, cells and faces can be treated as 'partially wet', and the impact of the sub-grid scale geometry may be represented by cells with larger sizes. The example below shows that a Quadtree model with 10/20m cell size (right) can convey water as smoothly as a 2.5/5/10/20m model along the narrow and meandering gully line.<br>
+
With the release of TUFLOW 2020, the combination of Quadtree mesh and Sub-grid Sampling (SGS) method has offered great flexibility for rapidly building flood models of varying resolution and increased topographic detail. Now a wide range of mesh sizes can be applied in a model based on the geographic feature and the area of interest in a project. Especially with the SGS method, cells and faces can be treated as 'partially wet', and the impact of the sub-grid scale geometry may be represented by cells with larger sizes.<br>
 +
The example below shows that a Quadtree model with 10/20m cell size (right) can convey water as smoothly as a 2.5/5/10/20m model along the narrow and meandering gully line.<br>
 
[[File:Fig1 H sgs.png|700px]]<br>
 
[[File:Fig1 H sgs.png|700px]]<br>
 
'''Figure 1 Water level simulation results with SGS. Left: 2.5/5/10/20m Quadtree model. Right: 10/20m Quadtree model.'''<br><br>
 
'''Figure 1 Water level simulation results with SGS. Left: 2.5/5/10/20m Quadtree model. Right: 10/20m Quadtree model.'''<br><br>
On the other hand, the same model running without SGS method creates notably different results in the distance of water travelled and the area of inundation on floodplain due to the poor geometry representation.<br>  
+
On the other hand, the same model running without SGS applied creates notably different results in the distance of water travelled and the area of inundation on the floodplain due to the poor topographic representation (one elevation per cell centre and cell face).<br>  
 
[[File:Fig2 H nonsgs.png|700px]]<br>
 
[[File:Fig2 H nonsgs.png|700px]]<br>
 
'''Figure 2 Water level simulation results without SGS. Left: 2.5/5/10/20m Quadtree model. Right: 10/20m Quadtree model.'''<br><br>
 
'''Figure 2 Water level simulation results without SGS. Left: 2.5/5/10/20m Quadtree model. Right: 10/20m Quadtree model.'''<br><br>
These examples indicate the mesh size sensitivities of a SGS model are significantly reduced, and thus modellers are increasingly using coarser mesh at areas far away from the area of interest. However, this has also create a challenge for how to interpolate water depth output at those coarser cells. The example below shows that even the 10/20m mesh and the 2.5/5/10/20m mesh SGS models produce similar water level, the depth output is much 'smoother' in the model with finer meshes. <br>
+
These examples show how the mesh size sensitivity are significantly reduced if using SGS. Therefore, modellers can confidently use a coarser mesh at areas away from the area of interest without adversely affecting the results. However, this has created a challenge on how to map results in areas of coarser mesh. For example, the images below show that whilst the 10/20m mesh and the 2.5/5/10/20m mesh SGS models produce similar water levels, the depth output is much 'smoother' in the 2.5/5/10/20m model due to the finer computational mesh. <br>
 
[[File:Fig3 D sgs zoom.png|700px]]<br>
 
[[File:Fig3 D sgs zoom.png|700px]]<br>
 
'''Figure 3 Water depth simulation results with SGS. Left: 2.5/5/10/20m Quadtree model. Right: 10/20m Quadtree model.'''<br><br>
 
'''Figure 3 Water depth simulation results with SGS. Left: 2.5/5/10/20m Quadtree model. Right: 10/20m Quadtree model.'''<br><br>

Revision as of 17:52, 1 June 2020

Introduction

With the release of TUFLOW 2020, the combination of Quadtree mesh and Sub-grid Sampling (SGS) method has offered great flexibility for rapidly building flood models of varying resolution and increased topographic detail. Now a wide range of mesh sizes can be applied in a model based on the geographic feature and the area of interest in a project. Especially with the SGS method, cells and faces can be treated as 'partially wet', and the impact of the sub-grid scale geometry may be represented by cells with larger sizes.
The example below shows that a Quadtree model with 10/20m cell size (right) can convey water as smoothly as a 2.5/5/10/20m model along the narrow and meandering gully line.
Fig1 H sgs.png
Figure 1 Water level simulation results with SGS. Left: 2.5/5/10/20m Quadtree model. Right: 10/20m Quadtree model.

On the other hand, the same model running without SGS applied creates notably different results in the distance of water travelled and the area of inundation on the floodplain due to the poor topographic representation (one elevation per cell centre and cell face).
Fig2 H nonsgs.png
Figure 2 Water level simulation results without SGS. Left: 2.5/5/10/20m Quadtree model. Right: 10/20m Quadtree model.

These examples show how the mesh size sensitivity are significantly reduced if using SGS. Therefore, modellers can confidently use a coarser mesh at areas away from the area of interest without adversely affecting the results. However, this has created a challenge on how to map results in areas of coarser mesh. For example, the images below show that whilst the 10/20m mesh and the 2.5/5/10/20m mesh SGS models produce similar water levels, the depth output is much 'smoother' in the 2.5/5/10/20m model due to the finer computational mesh.
Fig3 D sgs zoom.png
Figure 3 Water depth simulation results with SGS. Left: 2.5/5/10/20m Quadtree model. Right: 10/20m Quadtree model.

In TUFLOW, the depth output is interpolated from the depths at the nearest cell centre and cell corners surrounding the output grid. Because TUFLOW doesn't store the underlying high resolution DEM values at the moment, the interpolated depth may not perfectly represent the actual depth as illustrated in the figure below.
Fig4 sgs depth interporation.png
Figure 4 TIN interpolation used for water depth map output with SGS.

While we are developing high resolution SGS output in the future release to address this issue, we have also added a new functionality in the ASC_to_ASC utility to 'remap' a water level grid to a finer DEM grid. This page introduce how to use the ASC_to_ASC remap function, and also discuss the limitation of the method.

Remap Water Level to Finer DEM

The remap function in the ASC_to_ASC remaps a water level grid to a high resolution DEM using TIN interpolation. It outputs a high resolution water level grid and a high resolution depth grid.

asc_to_asc.exe -remap -wl lowres_h.asc -dem DEM_highres.asc

In this command:
"-wl <wl_file>"
Sets the coarser resolution water level grid to remap from.
"-dem <dem_file>"
Sets the finer resolution DEM.
The figure below overlays the remapped water depth on top of the 10/20m model water level output, and compared it with the original water depth output side by side. As can be seen, smooth water depth along the gully and the flood fringes is produced by the post process utility.
Fig5 D sgs remap zoom.png
Figure 5 Remapped vs original water depth for 10/20m mesh SGS model.

The next example presents the water level output and the remapped depth output in a river flood model. As can be seen, large portions of small buildings are submerged by water level output due to the relative size of the mesh and the buildings. However, remapped depth has clear dry/wet borders around the buildings.
Fig6 IR.png
Figure 6 Water level output and remapped water depth in a river flood model.

Remap Other Output Grids

The utility can also remap extra grid files (e.g. a hazard output) to the resolution of the DEM file.

asc_to_asc.exe -remap -wl lowres_h.asc -dem DEM_highres.asc lowers_hazard.asc

This command reads in an extra grid 'lowers_hazard.asc' and remaps it to the finder DEM resolution.The figure below compares the original and the remapped hazard outputs from the 10/20m SGS model.
Fig7 ZAEM1 sgs.png
Figure 7 Original vs remapped hazard output for 10/20m mesh SGS model.

It is important to note that, for any output types other than depth, this utility does NOT interpolate the result from the coarser grid to the finer grid, but only extended/reduced the output extent to the dry/wet extent. Therefore, the resolution of the remapped hazard above remains the same as the original output grid. The interpolation is not conducted for the following reasons:

  • Hazard categories are usually depended on both water depth and velocity, and it is not straight forward to interpolate a cell averaged velocity to sub-cell scale velocities with varying water depth.
  • Even if this can be done based on empirical relationship between depth and velocity, the obtained velocity is much less reliable than an actual simulation result from a model with finer mesh.

Therefore, we highly recommend refining the mesh size directly if users want to obtain finer velocity or hazard output at certain locations. The figure below shows the hazard output from the 2.5/5/10/20m model, and as can be seen the hazard result is much smoother along the gully.
SGS 02-20m ZAEM1.png
Figure 8 Original hazard output for 2.5/5/10/20m mesh SGS model.

Tips: multiple file names or wildcard are allowed for the extra grid files for remapping.

Model Mesh Size vs Remap Result

Beside the quality of hazard output discussed above, the model mesh size can also impact the remapped output at location with steep slope.

Road Crest

When water flows over a road crest, the remapped water depth may become negative if the output grid size is too coarse. The figure below shows the remapped depth over a road crest from models with mesh sizes ranging from 5m to 1.25m. As can be seen, some area doesn't have remapped depth despite the water is clearly over-topping the road.
Fig9 road mesh size.png
Figure 9 Remapped depth over a road crest from different mesh size models.

As illustrated in the next figure, the DEM has mush smaller resolution and the elevation changes rapidly at the road crest, when interpolation the coarser water level grids to the finer DEM, the interpolated water level may become lower than the local DEM level. In the 2D solver, this type of location is treated as 'weir flow' and proper depth is used for calculation. However, this is not reflected in the depth output yet and will be further addressed in the high resolution SGS output feature scheduled in the future release.
Fig10 road mesh size2.png
Figure 10 Modelled water level line over the road crest.

Upstream Catchment

In direct rainfall models, modellers can also benefit from applying relatively large mesh size and SGS method at the upstream region of a catchment. The SGS faces correctly capture the lowest elevations along the gullies to preserve the sub-cell scale flow paths, which improves the hydrologic response for a whole catchment model (see Kitts et. al., 2020, will add link to Duncan's paper in https://www.tuflow.com/Library.aspx when it is uploaded). However, if the cell size gets too large at the upstream region, only small portion of the cell becomes wet due to the large difference in elevation inside a cell. This makes the depth plotting extremely challenging and the remapped depth may become negative even though water is flowing through a valley. The blue contour below shows the remapped water depth of a 60m mesh whole catchment direct rainfall model. As can be seen the flow paths are not continuous along some valleys. The remapped water depth from a 20m mesh model is also plotted as the red contour underneath the 60m result. The flow path has becomes clearer as the mesh size is refined.
Fig11 Innisfall remapped D.png
Figure 11 Remapped depth at upstream catchment from different mesh size models.

Conclusion

The benefits of using the combination of Quadtree mesh and Sub-grid Sampling method are many. In this page we focused on the ability of representing the sub-grid scale geometry by SGS method, which allows user to apply coarser mesh to reduce the total simulation time. However, the interpolation of depth output of at coarse SGS mesh can be challenging as illustrated in the examples above. While we are developing high resolution output feature to tackle this issue, the remapping functions of the ASC_to_ASC utility can be used to post process water level outputs to high resolution depth results. This produces smooth depth output along main flow paths and flood fringes. However, SGS method and remapping tool is not a panacea, and computational mesh with sufficient resolution is still needed to produce reasonable and meaningful simulation results.

Finally, both SGS and remapping are relatively new features in TUFLOW and we are also learning from the users who apply them in their projects. If you find some results interesting and would like to share with the TUFLOW community, please feel free to email support@tuflow.com.