NRCS Curve Number Runoff
SCS / NRCS curve number method (TR-55) for computing runoff depth and volume from a design storm. The standard hydrologic-loss method for stormwater design across the US.
Defaults: 3.5 in (typical 25-year, 24-hr in mid-Atlantic), CN = 75 (suburban residential, B soils), 10-acre site.
Curve number guidance (TR-55 Table 2-2)
CN depends on land cover, hydrologic soil group (A/B/C/D), and antecedent runoff condition (ARC II is standard). Sample values for ARC II:
- Open space, good cover (parks, lawns > 75% grass): A=39, B=61, C=74, D=80
- Residential, 1/4-acre lots (38% impervious): A=61, B=75, C=83, D=87
- Residential, 1/2-acre lots (25% impervious): A=54, B=70, C=80, D=85
- Commercial / business (85% impervious): A=89, B=92, C=94, D=95
- Industrial (72% impervious): A=81, B=88, C=91, D=93
- Paved roads, curb & gutter: A=98, B=98, C=98, D=98
- Gravel roads: A=76, B=85, C=89, D=91
- Pasture / grassland, fair: A=49, B=69, C=79, D=84
- Forest, fair cover: A=36, B=60, C=73, D=79
- Row crops, contoured, good: A=64, B=75, C=82, D=85
For mixed-cover watersheds, compute area-weighted composite CN: CNw = Σ(CNi × Ai) / Σ(Ai). Same approach as composite C in the Rational Method.
Hydrologic soil group quick guide
- Group A: Sands and gravels with high infiltration. Sand, loamy sand, sandy loam.
- Group B: Moderate infiltration. Loam, silt loam.
- Group C: Slow infiltration when wet. Sandy clay loam.
- Group D: Very slow infiltration. Clay loam, silty clay loam, clay, or any soil with shallow bedrock.
Site-specific soil group from NRCS Web Soil Survey: websoilsurvey.nrcs.usda.gov. The default for design when soils data is missing is usually Group C or D depending on regulator.
Initial abstraction: 0.2 vs 0.05
The classic NRCS method assumes Ia = 0.2 S — a coefficient set in the 1950s based on a small data set. More recent analysis (Hawkins et al. 2002, NRCS NEH-630 ch. 10, 2017) found Ia/S is closer to 0.05 on most watersheds.
Switching from λ = 0.2 to λ = 0.05 increases runoff depth at low rainfall events significantly. For a 1-inch storm on CN = 75, λ = 0.2 gives Q = 0.04 in; λ = 0.05 gives Q = 0.31 in (8× higher). The two methods converge at higher rainfall depths.
Most US regulatory agencies still require λ = 0.2 for stormwater design. NRCS-published documentation generally uses 0.05 for current research. Use 0.2 unless you're certain your reviewer accepts 0.05.
From runoff depth to peak flow
This calculator gives total runoff depth and volume. To convert to a peak flow rate, you need a hydrograph method:
- TR-55 graphical peak discharge (Chapter 4): peak flow = qu × Am × Q × Fp, where qu is unit peak from the chart based on Tc and Ia/P.
- NRCS unit hydrograph (TR-20, HEC-HMS, HydroCAD): full hydrograph generation by routing rainfall through the dimensionless unit hydrograph.
- SWMM (more flexible): subbasin-by-subbasin routing with non-linear reservoir or kinematic wave.
For a quick check on small watersheds, the Rational Method calculator gives a peak flow without the depth/volume — but it doesn't match what NRCS produces because the assumptions differ.
Reference: USDA NRCS (1986). Urban Hydrology for Small Watersheds (TR-55). Hawkins, R.H., et al. (2002). "Runoff Curve Number Method: Examination of the Initial Abstraction Ratio." EWRI World Water and Environmental Resources Congress 2002.