Use the USDA NRCS SCS-CN (Curve Number) method to estimate a basin's direct runoff depth, runoff coefficient and peak flow from 24-hour rainfall depth. Vary CN, the antecedent moisture condition (AMC), basin area and initial abstraction ratio λ to support stormwater design, flood hazard mapping and green-infrastructure evaluation.
Parameters
24-hour rainfall depth P
mm
24-hour cumulative rainfall for the design storm (50-200 mm typical for urban drainage)
Blue raindrops fall; part is removed as initial abstraction (canopy and soil), the rest becomes surface runoff (blue flow) collected by the channel. Higher CN increases runoff, lower CN increases infiltration.
Q vs P curves (CN parameter sweep)
Rainfall breakdown — initial abstraction / infiltration / direct runoff
Conversions to AMC I/III (Hawkins 1985). Tabulated CN values are for AMC II; this formula adjusts CN for dry or wet antecedent soil-moisture conditions.
Runoff volume V and a 3× factor applied to the 24-hour average flow as a first-cut peak estimate. Detailed design should use a unit hydrograph or SCS triangular unit hydrograph method.
Rainfall-runoff estimation by the SCS-CN method
🙋
I hear about the "SCS-CN method" a lot in stormwater design. What does it actually do better than just "rainfall × runoff coefficient"?
🎓
Good question. A flat "rainfall × runoff coefficient" predicts the same fraction running off whether you get 5 mm or 500 mm — which is unphysical. SCS-CN fixes that with Q = (P − Ia)² / (P − Ia + S), a curve that gives almost zero runoff for light rain and asymptotically approaches the full rainfall once the soil saturates. Only one basin parameter — the Curve Number — is needed, so anyone with a land-use and soils map can compute it. USDA NRCS developed it in the 1950s and it is now the world's standard for stormwater and river planning.
🙋
How is the Curve Number actually picked? Just from a map and a table?
🎓
Yes — there is a table in USDA NEH-4 indexed by land use × "hydrologic soil group" (A through D). For example, lawn on group B soils is CN = 69, residential with 60% impervious on group C is CN = 90, and so on. If a basin has multiple land uses, you area-weight the CN. In Japanese urban drainage you commonly see CN = 85-90 for dense residential, 92-95 for commercial, and 70-80 after adding parks and green infrastructure.
🙋
For AMC I/II/III, which one should I use in design? Moving the slider really changes the answer.
🎓
It depends on the purpose. For design storms used to size sewers or detention basins, engineers typically pick AMC III (wet) as a conservative choice. For annual water-balance or green-infrastructure performance, AMC II is more representative. AMC I (dry) shows up in arid-zone or irrigation studies. Using Hawkins' conversions, CN = 75 at AMC III becomes CN ≈ 88 (runoff almost doubles); at AMC I it drops to CN ≈ 57 (runoff almost zero). Choosing the AMC is a real engineering judgment, balancing safety against over-design.
🙋
What is actually included in "initial abstraction Ia"? And where does λ = 0.2 come from?
🎓
Ia is the rainfall consumed before runoff starts, made up of (1) canopy interception (water held on leaves and stems), (2) depression storage (puddles), and (3) the initial infiltration that occurs before the soil reaches the point of runoff. The model assumes that runoff only begins after Ia = λ·S has been absorbed. The standard λ = 0.2 was derived empirically by USDA in 1954 from Midwestern US data and has been used ever since. More recently Hawkins and others have shown that λ = 0.05 actually fits measured data better, especially in urban basins, so modern practice is increasingly to use a smaller λ. Pick 0.05 to be conservative or 0.2 for tradition.
🙋
The "peak flow estimate" is just the average × 3, right? Can I use that directly for design?
🎓
Sharp eye — no, not for final design. Real design uses the SCS triangular unit hydrograph, the rational method, or a full rainfall-runoff model (HEC-HMS, RRI, SWMM). The "average × 3" in this tool is just an order-of-magnitude indicator; actual peak-to-average ratios depend on the time of concentration and can range from about 2 to 5. It is, however, a useful sanity check at the concept stage when you want to know roughly "how many m³/s does 100 mm produce in this catchment?". Use it to set direction, then run a proper hydrograph model for detailed design.
Frequently Asked Questions
CN is determined from the combination of land use and soil type (hydrologic soil group A-D). Typical values are 30-60 for forest and lawn, 60-80 for grassland and farmland, 80-95 for urban areas, and 98 for concrete pavement. In practice, engineers look up values in USDA NEH-4 tables and area-weight them across the basin. For example, a basin that is 60% residential (CN=85) and 40% park (CN=55) gives a weighted CN ≈ 73.
AMC (Antecedent Moisture Condition) is defined by the cumulative rainfall in the 5 days before the storm. AMC II is the average condition and uses the tabulated CN directly. AMC I represents dry conditions (5-day total < 13 mm) and reduces CN, increasing infiltration. AMC III represents wet conditions (> 28 mm) and increases CN. This tool uses the Hawkins (1985) conversions: CN_I = 4.2·CN/(10−0.058·CN), CN_III = 23·CN/(10+0.13·CN). Design flood calculations typically adopt the conservative AMC III, while annual water-balance studies use AMC II.
Initial abstraction Ia is the rainfall consumed by canopy interception, depression storage and initial infiltration before runoff begins, modeled as Ia = λ·S. λ = 0.2 was derived empirically from USDA data in the 1950s and remains the most widely used value. However, recent research (Hawkins et al.) shows that λ = 0.05 fits many measured datasets better, especially for urban basins. Smaller values are recommended for urban or arid basins, while larger values may suit forested basins. This tool lets you sweep 0.05-0.30 for sensitivity analysis.
The SCS-CN method gives daily direct runoff from a 24-hour rainfall total and loses accuracy when: (1) the storm is sub-hourly and peak intensity matters; (2) the basin is much larger than 100 km²; (3) snowmelt or frozen ground is involved; (4) winter or thaw seasons are simulated. For these cases use kinematic-wave models, HRU-based distributed models, or radar-coupled rainfall-runoff models (RRI, SWAT). For preliminary design and screening of small-to-mid basins, SCS-CN remains very useful.
Real-World Applications
Urban stormwater and detention-basin design: SCS-CN is the global standard for residential and commercial development approvals, storm-sewer sizing and detention-basin volume calculations. For example, a 10-hectare site changing from CN = 65 (grass) before development to CN = 88 (residential) after development sees its 100 mm direct runoff jump from about 30 mm to 70 mm — more than doubling. The detention-basin volume that absorbs that difference is the heart of drainage planning, and this tool lets you switch "pre" and "post" to compare runoff volumes instantly.
Flood hazard mapping and risk assessment: Municipal hazard maps use SCS-CN to estimate direct runoff from 100-year rainfalls (e.g. 300 mm/24h), feed it into channel-routing models, and delineate flood-inundation zones. The method is also used to forecast how much future floods will worsen under land-use change — deforestation, conversion of farmland, urbanization — and is a standard input to municipal planning documents across the Tokyo, Nagoya and Osaka metropolitan regions and similar urban areas worldwide.
Green-infrastructure performance evaluation: Rain gardens, permeable pavements, green roofs, and bioswales can all be quantified as CN reductions. Replacing 1 ha of parking lot (CN = 98) with permeable pavement (CN = 80) reduces 100 mm direct runoff from 84 mm to 47 mm. Multiply by area to obtain a volumetric saving, combine with CO₂ and wastewater-treatment savings, and you have a defensible cost-benefit story — exactly the pattern that has become standard in recent green-infrastructure investment cases.
Agricultural water-balance and erosion modeling: SCS-CN is the core surface-runoff module of distributed hydrological models such as SWAT (Soil and Water Assessment Tool) and HEC-HMS. These models simulate long-term water balance (ET, runoff, soil moisture) for agricultural basins and the wash-off of sediment and nutrients from farmland. Researchers and consultants use them to quantify how rotational cropping or best-management practices (BMPs) change downstream flow and pollutant loads.
Common Misconceptions and Pitfalls
The biggest pitfall is simply averaging CN values across an inhomogeneous basin. If a basin is half pavement (CN = 98) and half forest (CN = 30), naively averaging gives CN = 64, but this is not physically correct: because S depends nonlinearly on CN, you should compute Q separately for each land use and area-weight the resulting Q. Pragmatically, area-weighted CN is acceptable when the spread is small, but when CN values differ by 20 or more, the basin should be split. Watersheds mixing permeable and traditional pavements, or farmland and forest, deserve special care.
Second, confusing peak flow with direct runoff depth. SCS-CN outputs Q (direct runoff depth in mm), which is the total water leaving the basin in 24 hours divided by area — no time information is included. Converting to a channel peak flow (m³/s) requires assuming a temporal distribution or using a unit hydrograph. The "× 3" peak in this tool is a very coarse first-cut value; for embankment, bridge or culvert design always use HEC-HMS, the rational method, or the SCS triangular unit hydrograph to model the time response properly.
Finally, treating NEH-4 CN values as universal truth. NEH-4 values were calibrated on US Midwestern test basins from the 1950s-60s and do not always fit Japanese baiu/typhoon storms, volcanic-ash (kuroboku) soils, clear-cut areas, terraced rice fields, and similar regional conditions. When historical runoff records exist, back-calculate CN from observed runoff (basin calibration). When they do not, run sensitivity analyses across AMC and λ to show the band of plausible outcomes — that is the honest engineering practice. The "storage correction factor" in this tool serves as a stand-in for that local calibration.
How to Use
Enter rainfall depth in mm (e.g., 85 mm for a 24-hour storm event)
Input Curve Number (CN) between 30–100 based on soil group and land use; typical urban areas range CN 70–85, pasture CN 58–72
Specify basin area in hectares or km²; adjust initial abstraction ratio if site-specific infiltration data available (default 0.2)
Select appropriate Antecedent Moisture Condition (AMC I, II, or III) to adjust CN for soil wetness conditions
The simulator calculates maximum retention S, initial abstraction Ia, direct runoff Q, and peak flow using unit hydrograph routing
Worked Example
For a 156 ha suburban catchment in Ohio receiving 75 mm rainfall: CN = 75 (composite for 60% residential, 40% grassland), AMC II applied. Adjusted CN = 75. Maximum retention S = (25400/CN) – 254 = 83.7 mm. Initial abstraction Ia = 0.2 × 83.7 = 16.7 mm. Direct runoff Q = (75 – 16.7)² / (75 + 83.7 – 2 × 16.7) = 28.4 mm. Peak flow estimate (rational method with concentration time 2.1 hours) = 156 × 28.4 / 3.6 / 2.1 ≈ 5.8 m³/s discharge at outlet.
Practical Notes
Increase CN by 1–3 points for AMC III (saturated antecedent conditions typical in spring) to account for reduced infiltration capacity during wet seasons
For impervious urban surfaces (roofs, pavement >80% coverage), use CN 85–90; pervious zones (forests, undisturbed grassland) use CN 45–60
Validate peak flow estimates against USGS streamflow gauges if available; SCS method often underestimates peaks in flashy urban channels with steep slopes (>5%)
Initial abstraction ratio varies 0.05–0.3 depending on surface storage; use 0.3 for depressional areas like stormwater ponds