The choice of a baseflow method hinges upon the purposes of the study and desired level of detail.  Typically, the amount of baseflow generated during extreme events (like those simulated within dam safety studies) pales in comparison to the amount of direct runoff.  However, in locations with voluminous underground flow paths, direct runoff may not be the predominant source of streamflow.  Also, studies that require the simulation of short-duration events (e.g., days to weeks) tend to focus more upon direct runoff generation while studies that require the simulation of long-duration events (e.g., months to years) use continuous simulation techniques that focus on both direct runoff and baseflow generation.

Initial estimates of baseflow parameters should be subjected to a model calibration process where computed outputs are compared against observed data and model parameters are modified in order to achieve an adequate fit.  Also, best estimate parameters derived through the aforementioned model calibration process should be tested through a model validation process where computed results, without any further parameter modifications, are used to compute outputs which are compared against observed data for independent events that were not considered during model calibration.  Storm events used during both model calibration and validation should be approximately equal to the magnitude of events that are being considered within the particular application.

The following table contains a list of various advantages and disadvantages regarding the aforementioned baseflow methods available for use within HEC-HMS.  However, these are only guidelines and should be supplemented by knowledge of, and experience with, the methods and the watershed in question.

Method

Advantages

Disadvantages

Constant Monthly

  • Simple, parsimonious method.
  • Method may be too simple to predict baseflow especially during large storm events.
  • Does not automatically conserve mass.
  • No connection to infiltrated water.

Recession

  • "Mature" method that has been used successfully in thousands of studies throughout the U.S.
  • Parameters can be regionalized.
  • Baseflow can be reset after a storm event.
  • Does not automatically conserve mass.
  • No connection to infiltrated water.

Bounded Recession

  • Similar to advantages of the Recession method.
  • Similar to disadvantages of the Recession method.
  • Baseflow cannot be reset after a storm event.
  • Not well documented nor widely used.

Linear Reservoir

  • This method automatically conserves mass.
  • Method is scalable; can be used for single event or continuous simulation.
  • Infiltrated water can be partitioned to volume that returns to the stream and volume that is lost to deep aquifer storage.
  • Baseflow generation corresponds to infiltration (more infiltration = more baseflow; less infiltration = less baseflow).
  • Parameters can be regionalized and estimated using watershed characteristics.
  • Requires more parameters than the Recession method.
  • Less parsimonious than simpler, empirical methods.

Nonlinear Boussinesq

  • Similar to advantages of the Recession method.
  • Parameters can be estimated using watershed characteristics.
  • Similar to disadvantages of the Recession method.
  • Requires more parameters than the Recession method.
  • Not well documented nor widely used.