Multiple factors limit the precision and certainty of the model results.
- This model is based on the pure translation method, which assumes that all flow at any particular location flows with the same speed. The model does not take into account storage effects of the watershed. Thus, there is no flow spreading and all flow elements starting from a particular cell reach the watershed outlet simultaneously. This method is only applicable to small watersheds with minimal storage effects.
- The unit hydrograph is a linear response function of the watershed. It assumes that the time base of the hydrograph remains constant regardless of the amount of runoff resulting from different storms with the same duration. Therefore, runoff response from a storm with a runoff depth other than one unit can be obtained by multiplying the runoff depth by the ordinates of the unit hydrograph developed for that duration.
- The methods described for deriving a spatially distributed velocity field rely on a number of assumptions. The velocity field is spatially variant but time and discharge invariant. The velocity field used is dependent on local constant variables such as local slope and upstream contributing area (flow accumulation) and not on time-varying variables such as flow or storage.
- This model uses an approach similar to the one proposed by Clark (1945).[2] In Clark's model, flow is routed through a pure translation channel followed by a single reservoir for the whole watershed. However, the model used in these tutorials is based on pure translation and does not account for any storage effects. Another difference is that this model allows subareas within the watershed, down to the cell size scale, to be considered as separate units, each with its own distinct response function at the watershed outlet. Additionally, this model's framework has the flexibility to allow precipitation to vary by isochrone zones while estimating direct runoff hydrographs.
