About the model

citySponge estimates, for every one of New York's ~111,000 street segments, what happens to rain at the peak of a steady storm: how much lands on the block and its lots, how much runs along the curb, how much the catch basins take in, whether the sewer pipe below can carry it, and where it ends up: a treatment plant, a river through a combined sewer overflow, or a puddle at the bottom of the hill.

How it works

Peak flows use the Rational method (Q = C·i·A) with runoff from the street, sidewalks and the lots on each side. Gutter flow uses Izzard's equation to find spread and depth at the curb. Each block's inlets capture up to their capacity; the rest flows downhill to the next block or ponds at a low point. Pipe capacity is sized from a design storm on an NOAA Atlas 14 rainfall curve for Central Park. At combined sewer regulators, the interceptor carries a multiple of dry-weather flow to the treatment plant and the rest overflows.

Data

• NYC Open Data: DoITT street centerline (CSCL), DEP catch basins, DEP outfalls, DEP MS4 drainage areas.
• Ground elevation: USGS 3DEP via AWS Terrain Tiles.
• 311 flooding complaints: NYC Open Data 311 service requests, where available.
• Basemap: CARTO, OpenStreetMap contributors.

What to keep in mind

• The sewer network is a proxy. DEP's pipe network is not public. Pipes here are routed along streets to the outfall that serves each area, with uphill routes penalised. Real pipes, sizes and regulators differ.
• Steady-state peaks. Flows are Rational-method peaks for a storm of uniform intensity, not a time-varying simulation. Storage in pipes and tidal backwater are ignored. Storm volumes assume the peak lasts the whole storm, so they run high.
• Elevation is coarse. A ~10 m terrain model misses curb cuts, small dips and building drainage.
• Not a substitute for DEP's Stormwater Flood Maps, FEMA maps, or local knowledge. Use it to explore how the system works, not to judge a specific property.