Mesh Regions

Mesh regions let you vary the computational mesh resolution across your domain. Use a finer mesh in the areas of interest and a coarser one elsewhere, to balance accuracy against runtime and cost.

Properties

Property	Type	Description
resolution	number	Target triangle edge length in metres inside this polygon

How they work

Each mesh region is a polygon with its own target resolution. At build time the regions are passed to ANUGA's create_mesh_from_regions as interior regions, which constrains the maximum triangle area inside each polygon while the rest of the domain keeps the scenario's base mesh size.

Resolution is an edge length, so halving it roughly quadruples the triangle count inside that polygon.

A region has to be attached to the scenario

Drawing a mesh region does not apply it. The scenario has to reference the mesh regions layer, and Hydrata warns you when a region is drawn but not attached: it will be ignored at build time and the mesh will come out uniform instead.

That warning is worth reading, because a uniform mesh still builds and still runs. Nothing fails; you just get the coarse answer you did not ask for.

What it costs

Triangle count drives build time, run time and price, and mesh regions are usually what drives triangle count.

The scenario shows an estimated triangle count before the build and the actual count after it, so you can see what a region really did. When the regions are the dominant cost, Hydrata says so, along with the share they account for, so you do not spend time tuning the base mesh size when the base mesh size is not the problem. Structure hole boundaries can drive the count the same way.

Tips

• Use a coarse base mesh size for the full domain, for example 20 m
• Add mesh regions at 2 m to 5 m where you need detailed results
• Check the built triangle count against the estimate before scaling up
• A finer mesh significantly increases triangle count, runtime and price

See Configuring Runs for the base mesh size and what the build reports.