“intuitive ways of physical modelling with data-driven design strategies”

Hermansdorfer et al., Journal of Digital Landscape Architecture, 2020
FIELD LAB / 01

One site. Three questions.

Follow rain downhill, turn the sun, or explore wind around a courtyard building.

192 M × 192 M / CONTOURS EVERY 4 M

Move over the map or tap to inspect a point.

How this model works

Runoff uses steepest descent to one of eight neighbours (D8) on a priority-flood spill surface, then sums upstream effective rainfall. Width indicates discharge in litres/second; roof cells are impermeable and tree cover increases infiltration. Small depressions fill to their spill level. Traces use continuous distance along complete flow paths; their playback speed is illustrative. Sunlight uses surface orientation and ray-traced terrain, building and canopy shadows; white is stronger direct exposure, black is shade. Wind uses a D2Q9 lattice-Boltzmann solver with no-slip building walls and simplified canopy drag. Wake circulation develops from the evolving flow. This is an illustrative 2D slice with simplified boundary conditions, not atmospheric turbulence or a calibrated comfort assessment. These are synthetic studies, not site predictions.

D8 method ↗ · Solar incidence ↗ · Lattice-Boltzmann method ↗

The map as an interface

Urban planning decisions are made across overlapping systems. A low point in terrain can become a stormwater route; a tree canopy changes shade, microclimate and habitat; wind exposure affects public space; roof orientation changes solar potential. Layering these views helps a workshop group ask better questions than a single polished plan can answer.

In the Sandscape and Datascape paper, a physical sand surface and augmented projection form a tangible landscape interface. Participants reshape terrain while digital feedback responds. Bridging Tangible and Virtual Realities extends this idea toward participatory procedures in a visual programming environment. Depth sensing, projected topography and cut/fill analysis can make a change visible at the moment a participant moves sand.

I have also explored robotic sand manipulation with Aalto collaborators. The physical table and the digital model can meet in a loop: scan, interpret, project, discuss, modify. Watersheds and stormwater, solar energy, forestry data, microclimate and wind analysis can all enter that conversation as layers. The illustration on this page is an interactive conceptual map of that approach, not a measured site study.

From data to a shared decision

A planning tool succeeds when participants can see which layer produced a claim and which assumptions sit behind it. The aim is to let specialists bring rigorous data into a room without making everyone else a passive audience.

REFERENCES & FURTHER READING

Short paper excerpts above are attributed to their authors. Commentary is Kane Borg’s interpretation.