Variable Clark Unit Hydrograph Parameter Regression Equations For California

The Clark unit hydrograph transform (Clark, 1945) is the most commonly employed runoff transform method within hydrologic modeling applications undertaken by the U.S. Army Corps of Engineers (USACE). This method is mature, well established, well documented, and simple to set up and use. Also, parameters can be regionalized, related to measurable basin characteristics, and varied with excess-precipitation rates, which are important traits for methods used in dam safety studies.

The new Variable Clark unit hydrograph method contained within the Hydrologic Engineering Center’s (HEC) Hydrologic Modeling System (HEC-HMS) builds upon the “classical” Clark unit hydrograph method by allowing parameters (i.e. time of concentration and storage coefficient) to change throughout a simulation. This new method allows modelers to efficiently simulate the dynamic runoff response of a watershed when subjected to large excess precipitation rates, such as those expected during design storms like the Probable Maximum Precipitation (PMP). The California Department of Water Resources, Division of Safety of Dams (DSOD) partnered with HEC to develop methods to estimate Variable Clark unit hydrograph parameters throughout the state of California for use within dam safety applications. Through the use of these equations, synthetic unit hydrograph parameters can be quickly derived for independent watersheds when observed data is not readily available.

New tools within HEC-HMS were used to construct numerous hydrologic models, develop initial parameter estimates, and calibrate parameters using observed data. Regression equations that relate physically measurable watershed characteristics (e.g. longest flow path) to Variable Clark unit hydrograph parameters (e.g. time of concentration and storage coefficient vs. excess precipitation rates) were then developed. Predicted Variable Clark unit hydrograph parameters obtained through the use of the aforementioned regression equations were then evaluated using a validation process. The results of this validation process demonstrated that each regression equation adequately predicted Variable Clark unit hydrograph parameters throughout the state of California.