Guides
Each guide explains what a plate computes and what it does not. Read them before you use the tool, or when a plate looks odd and you want to know why. They contain the numbers we actually measured and the traps we actually fell into.
Getting started
- How to draw a site analysis diagram
Order and principles, and which parts automation takes over.
- The panels — what to read from each
Why the default set is composed the way it is; the questions each plate answers.
- Site overview and section — zoning limits and a 1:1 skyline cutnew
Zoning by share of the site, the FAR and height limits each city's data carry, max floor area, the area around in numbers, and a section through the site at true vertical scale.
- Frequently asked questionsnew
Missing buildings, gray unknowns, radius, compute times, files, SketchUp, licensing.
Sun, wind and heat
- Sun path and shadow diagrams in the browser
How the sun-path diagram works, the 3D shadow model, changing date and hour.
- Sun hours map — a whole day of shadow in one platenew
Shadows accumulated every ten minutes. True solar time, the criteria shown, the design we discarded, and why it is not a shadow study.
- Wind roses and urban airflow diagrams
Sixteen-sector observed frequencies and airflow between buildings (LBM). Grid, domain, and how it differs from a wind assessment.
- Thermal comfort (UTCI) mapnew
Weather, wind, shade and ground cover folded into a felt temperature. The 26 ft canopy assumption and the known biases.
Ground and trees
- Slope diagram — reading percent gradenew
Bands at 5·10·15·25%, the 164 ft smoothing and what it hides, and why it is not a survey.
- Plot size diagram — urban grain from tax lotsnew
Lot area in five bands. Where US parcels come from, the count-versus-area trap, the ownership caveat.
- Finding trees in aerial imagerynew
NAIP near-infrared, NDVI, the offset that turned the Hudson into forest, and recall checks against the street tree records of New York (86%), Los Angeles (84.5%), San Francisco (89.2%) and Boston (86.4%).
Export and finishing
- Finishing the SVG in Illustrator
Layer structure, recoloring, layout and print resolution, DXF.
- Exporting the 3D context model — GLB, 3DM, SKPnew
Format differences, how the color mode carries into the file, opening in SketchUp, and which buildings are missing.
Cities
- New York City
1,091,427 building footprints, 98.6% with year built — sample diagrams and the sources used there.
- Los Angeles
1,097,475 building footprints, 96.0% with year built — sample diagrams and the sources used there.
- San Francisco
162,899 building footprints, 96.2% with year built — sample diagrams and the sources used there.
- Boston
105,587 building footprints, 96.3% with year built — sample diagrams and the sources used there.
- Washington, DC
171,784 building footprints, 94.2% with year built — sample diagrams and the sources used there.
- Chicago
832,475 building footprints, 96.9% with year built — sample diagrams and the sources used there.
- Miami
72,945 building footprints, 96.2% with year built — sample diagrams and the sources used there.
- Dallas
334,650 building footprints, 94.8% with year built — sample diagrams and the sources used there.
- Houston
636,576 building footprints, 94.6% with year built — sample diagrams and the sources used there.
- Denver
266,329 building footprints, 98.5% with year built — sample diagrams and the sources used there.
- Seattle
300,350 building footprints, 97.1% with year built — sample diagrams and the sources used there.
- Minneapolis
177,282 building footprints, 99.1% with year built — sample diagrams and the sources used there.
US datasets
- Data sources
Every dataset behind the US edition — New York City, Los Angeles, San Francisco, Boston, Washington, DC, Chicago, Miami, Dallas, Houston, Denver, Seattle and Minneapolis — its publisher, and measured coverage.
- NYC MapPLUTO without GIS
Tax-lot polygons with year built and land use, and how to use them.
- LA County parcels without GISnew
The Assessor's 2.4 million parcels — use, year built, floor area — and the condominium stacks that break naive joins.