A unit hydrograph is the hydrograph of direct runoff produced at the outlet of a drainage basin by a unit depth of effective rainfall, distributed uniformly across the basin at a constant rate during a specified period. In hydrology, it provides a practical way to transform rainfall excess into a time-varying discharge. The unit depth is commonly 1 millimetre, 1 centimetre, or 1 inch, depending on the convention used. A unit hydrograph must therefore be identified by its basin, rainfall duration, and unit-depth convention.
Physical meaning and terminology
A storm hydrograph includes water arriving at a basin outlet through pathways with different travel times and storage effects. Some water reaches channels quickly, while other contributions arrive later. The unit hydrograph represents this combined response without explicitly calculating every flow path.
Its input is not total precipitation, but effective rainfall, also called rainfall excess: the portion assigned to direct runoff after accounting for interception, infiltration, and other losses or storage changes. Effective rainfall is usually estimated by a separate loss model.
Its output is direct runoff, rather than total streamflow. Baseflow, generally associated with slower subsurface contributions, is separated from an observed hydrograph before a unit hydrograph is derived. In applications, an independently estimated baseflow hydrograph can be added to the calculated direct runoff. Direct runoff is an operational modeling category and need not consist exclusively of water flowing over the land surface.
The duration of a unit hydrograph refers to the duration of its effective-rainfall input, not the duration of the resulting runoff. A one-hour unit hydrograph, for example, describes the response to one hour of uniform effective rainfall; discharge may continue for many hours afterward.
Important characteristics include:
- Peak discharge: the maximum direct-runoff rate per specified unit rainfall depth.
- Time to peak: the interval from the beginning of effective rainfall to peak discharge, unless another reference point is explicitly stated.
- Basin lag: commonly the interval between the centroid of effective rainfall and the hydrograph peak, although definitions vary.
- Time base: the interval over which the direct-runoff response is represented as nonzero.
Assumptions
Classical unit-hydrograph theory treats the basin approximately as a linear, time-invariant system.
Proportionality. For effective rainfall of a fixed duration and spatial distribution, multiplying the rainfall depth by a factor multiplies every direct-runoff ordinate by the same factor. Hydrograph timing and shape remain unchanged.
Superposition. Responses to successive rainfall pulses can be shifted in time and added. The runoff generated by a complex storm is consequently represented as the sum of responses to its component pulses.
Time invariance. A rainfall pulse produces the same response whenever it occurs, provided the relevant basin conditions remain unchanged.
Uniform input. The elementary rainfall pulse is assumed to have a constant effective intensity during its specified duration and a uniform spatial distribution over the basin.
These assumptions concern the transformation of effective rainfall into direct runoff. They do not imply that the relationship between total rainfall and runoff is linear. Rainfall losses may vary strongly with soil moisture, storm intensity, season, and previous rainfall.
Mathematical formulation
Let be the discharge hydrograph generated by a reference effective-rainfall depth occurring over duration . If basin area is , conservation of volume requires
with consistent units. Thus, one millimetre of effective rainfall over a basin of produces of direct runoff, regardless of the hydrograph’s shape.
Divide a storm into intervals of duration , with effective-rainfall depths . The predicted direct-runoff discharge is
where for negative time. Each term is a scaled, delayed copy of the same unit hydrograph. This operation is a discrete form of convolution.
For example, suppose a one-hour unit hydrograph is defined for 1 millimetre of effective rainfall. A storm producing 3 millimetres in its first hour and 2 millimetres in its second hour gives
The two responses overlap. Their combined peak depends on the unit hydrograph’s shape and the timing of the inputs; it is not generally the sum of their separate peak discharges.
Unit-hydrograph ordinates are often tabulated as discharge per unit rainfall depth. Under that convention, the reference depth is absorbed into the tabulated values. Distinguishing this convention from a hydrograph defined for an explicit depth prevents unit-conversion errors.
Derivation from observations
An observed unit hydrograph is commonly derived from a relatively isolated storm with adequate rainfall and streamflow records.
- Assemble observations. Estimate basin-average rainfall and obtain discharge measurements, typically from a stream gauge. Where discharge is calculated from water level, uncertainty in the rating curve also affects the result.
- Separate baseflow. Estimate the direct-runoff component of the event hydrograph.
- Calculate direct-runoff volume. Integrate direct-runoff discharge over the event.
- Determine effective-rainfall depth. Divide runoff volume by basin area and establish a rainfall-excess sequence consistent with the event.
- Normalize the hydrograph. For an event approximating a single uniform pulse, divide direct-runoff ordinates by the effective-rainfall depth expressed in reference-depth units.
- Compare events. Examine several suitable storms to assess whether a representative response can be identified.
Complex storms require an inverse calculation rather than simple division. Unit-hydrograph ordinates can be estimated from the rainfall–runoff convolution equations using methods such as least squares. Because noisy observations can yield oscillating or negative ordinates, practical estimation may impose smoothness, nonnegativity, and volume constraints.
Results depend on rainfall sampling, loss estimation, baseflow separation, and discharge accuracy. A fitted unit hydrograph can partly compensate for errors in those inputs, so agreement with a calibration event does not establish that the inferred response is unique.
Forms and related methods
Duration-specific unit hydrographs
A finite-duration unit hydrograph applies to a particular effective-rainfall duration. Changing that duration generally changes both peak discharge and shape.
For an integer multiple of the original duration, appropriately delayed copies of the shorter-duration hydrograph can be averaged to obtain a longer-duration unit hydrograph. More general duration conversion can use an S-hydrograph, formed by summing an indefinitely repeated sequence of unit hydrographs. Differencing two time-shifted S-hydrographs and rescaling their difference produces a response for another duration.
Instantaneous unit hydrograph
An instantaneous unit hydrograph is the idealized response to a unit depth of effective rainfall applied in an infinitesimally short interval. It is a mathematical impulse response, not a physically attainable storm. Finite-duration unit hydrographs can be obtained by averaging delayed instantaneous responses over the rainfall duration.
Synthetic and dimensionless unit hydrographs
A synthetic unit hydrograph is estimated from basin characteristics or regional relationships rather than derived solely from a suitable observed storm. Common approaches include Snyder’s method, the Soil Conservation Service/Natural Resources Conservation Service method, and Clark’s method.
Their structures differ. Snyder’s method uses regional relationships for hydrograph characteristics. The SCS/NRCS approach commonly scales a dimensionless hydrograph using runoff volume and timing parameters. Clark’s method combines a representation of travel-time distribution with linear-reservoir storage.
A dimensionless unit hydrograph expresses discharge and time as ratios to characteristic values, commonly peak discharge and time to peak. It supplies a standardized shape that can be scaled for a particular basin. Such shapes are empirical approximations, not universal basin responses.
Digital elevation models and geographic information systems support synthetic and spatially distributed approaches by providing basin boundaries, slopes, channel lengths, and estimates of travel-time distributions.
Historical development and applications
The American engineer Leroy K. Sherman introduced the unit-hydrograph method in 1932. It offered a systematic way to derive storm-runoff hydrographs from rainfall and observed basin response without resolving all of the underlying hydraulic processes.
Applications include flood-hydrograph estimation, reservoir inflow modeling, drainage design, flood forecasting, and event-based watershed simulation. Unlike the rational method, which conventionally estimates peak discharge, a unit-hydrograph model produces a full discharge time series. This allows calculation of flood volume, hydrograph overlap, and storage requirements.
In larger watershed models, separate subbasin unit hydrographs may generate local runoff hydrographs, which are then routed through channels and reservoirs. Runoff transformation and downstream routing are distinct operations, although an outlet-calibrated unit hydrograph already embodies the aggregate travel and storage effects within its own basin.
Limitations and uncertainty
Real basins are not strictly linear or time invariant. As discharge increases, flow velocities, inundated areas, channel storage, and contributing runoff areas can change. A hydrograph derived from moderate events may therefore represent extreme floods poorly.
Spatial rainfall patterns also matter. Two storms with the same basin-average effective rainfall can produce different hydrographs if their rainfall is concentrated in different tributaries or moves across the basin in different directions. This limitation becomes especially important for large basins or localized intense storms.
Antecedent wetness can affect both rainfall losses and runoff pathways. Seasonal vegetation, frozen soil, snowmelt, wildfire, and urbanization may alter the basin response. Changes in drainage infrastructure or land cover can make a previously calibrated unit hydrograph unrepresentative.
Temporal resolution is another constraint. A coarse rainfall interval can obscure short bursts that control a flash flood peak, whereas very fine intervals can demand detail unsupported by available observations. Model evaluation consequently considers peak magnitude, timing, volume, and performance across independent events, rather than relying on a single measure of fit.