Delfhsaven, Rotterdam, The Netherlands

Collect the necessary data

Step 1: Point clouds

Luckily the Netherlands has really high quality point clouds freely available. The format for point clouds can be multiple: *.laz, *.txt, *.ply. To get the point clouds there are two options now:

• Option 1: https://app.pdok.nl/rws/ahn3/download-page/

• Option 2: https://geotiles.nl/

In the case of Delfshaven these are the two tiles selected:

Step 2: Footprints

There are multiple footprints we need to get. First buildings footprints are necessary to extrude them, then we need semantic surfaces for water, vegetation and maybe other type of surfaces' materials of interest.

For buildings footprints we can go here:

• https://www.pdok.nl/introductie/-/article/basisregistratie-adressen-en-gebouwen-ba-1
• https:https://service.pdok.nl/lv/bag/wfs/v2_0

For other types of surfaces:

• https://app.pdok.nl/lv/bgt/download-viewer/
• Download "gmllight" type of data for the tiles of interest.

Reconstruct geometry

Set-up visualization and CRS

In order to prepare all this data together to use in City4CFD we need to run a set of preprocessing steps within QGIS:

• Open a map so you have proper visualization of the location of interest:

• Click on the coordinate system and change it to 28992 (local Amesfort system).

Import footprint polygons:

To import the buildings:

• Make a connection to 3DBAG:

• Then select pand on the left panel:

To import the other surfaces:

• From the data downloaded for other surfaces import for water and vegetation these two: waterdeel and begroeidterreindeel (see left panel).

Sample a limited area for your simulation set-up:

• Then we can sample using a scratch layer polygon in order to reduce the amount of data:

• Important, if there is not CRS information for the water and vegetation you need to add it with double click on the layers. Otherwise they might not be sampled at all by the polygon.

• To check the leftmost southmost point and rightmost topmost point in your domain I used the coordinate capture from QGIS and the polygon with which I sampled the data.

Merge polygons where necessary

• When you have troubles with the water surface, fix it by merging the polygons.

• Toggle editing, merge selected features.

• You can also dissolve several polygons, but be careful -- City4CFD will remove polygons that are bigger than the domain. If your merged polygons extend further than the domain boundaries, they will not be reconstructed. To mitigate the issue, you can split polygons at the boundaries of your domain.