Flood Inundation Simulation

Real-Time Watershed Hydrology

A storm hits a small watershed. SCS Curve Number converts rainfall to runoff. Water accumulates in low-lying terrain. Sensors go offline as the inundation rises. Press Run Storm and watch the basin fill.

PRECIP · 75 mm / 6 hr

T+0:00 · stage 0.00 m

SENSORS · ALL DRY

Storm Parameters

Precipitation 75 mm

Storm Duration 6 hr

Antecedent Moisture AMC II (avg)

Land Cover Pasture

Forest

Pasture

Devel.

Urban

Deep

Shallow

High ground

SCS Curve Number —

Potential retention S —

Runoff depth (Q) —

Inundated area —

Peak stage —

Watershed Sensors

SCS Curve Number Runoff

// Runoff depth (mm) Q = (P − 0.2·S)² / (P + 0.8·S) for P > 0.2·S S = 25400/CN − 254 // mm // CN: 55 forest · 74 pasture // 86 developed · 92 urban // Bathtub fill on DEM until // runoff volume balances

What the API call looks like

POST https://api.geog.ai/v1/simulate/flood { "watershed_id": "ws_colorado_basin_seg_42", "precip_mm": 75, "duration_hr": 6, "antecedent_moisture": "AMC_II", "model": "scs_cn" // scs_cn | swmm } // Response includes: // inundation Polygon (GeoJSON, time-stepped) // per-device predicted_depth_m + time_to_inundation_min // per-channel-reach predicted_stage_m // confidence interval (calibrated against gauges)

This demo runs entirely in your browser using SCS Curve Number runoff over a synthetic DEM — no API calls made. The production geog.ai flood engine adds USGS DEM tile lookups, EPA SWMM urban network routing, Muskingum-Cunge channel routing, and a calibration loop that refines curve numbers and Manning's roughness from gauge observations.