BANKFULL DISCHARGE

Bankfull Discharge is said to be the discharge which shapes and maintains the channel. It is the discharge that, over time, moves the most sediment and water. Being able to estimate and calculate bankfull discharge is a critical step in designing a successful restoration project. Bankfull discharge is defined as the discharge which just begins to overtop the flood plain. In other words identifying bankfull stage in the field is a matter of recognizing a flood plain. A flood plain is defined as a flat depositional surface adjacent to and being formed by the stream in its present hydrologic state.

Recognizing bankfull stage in the field is not necessarily trivial or simple. Not all stream types have well developed flood plains. When dealing with stable C channels, identifying the flood plain can be fairly straightforward. The tops of point bars are very distinct and reliable indicators of bankfull stage or the elevation of the flood plain. In B channels or A channels things may not be so straightforward. This is because B and A channels do not have well developed flood plains. Even in stable channels with no accelerated adjustment going on, only small or partial indicators of the flood plain may exist. However, there will be flat depositional features at various places along the channel, and they can be found.

The presence of terraces (abandoned flood plains) can also add confusion to the process of identification of bankfull stage. In regions where glaciation has occurred very distinct and persistent terraces may be present from past climatic regimes. For example in the rocky mountain regions there are three terraces from the Holocene period. There is a high terrace often 30 or more feet above the current flood plain. There is also a middle terrace 8-10 feet above the current flood plain. Finally there is a low terrace often 4 feet above the current flood plain. In the middle Atlantic region these terraces do not exist because glaciation did not occur in these regions. However streams in the middle Atlantic region have their own sources of confusion in developed or developing areas. Many streams especially smaller ones have experienced large changes in hydrologic regime as a result of land use changes which increase the quantity and frequency of runoff events in the watershed. A common adjustment pattern for these streams is incision or down cutting. When this occurs, these streams leave terraces as flat depositional surfaces which are higher than the current bankfull stage. If a terrace is mis-identified as a flood plain, overestimates of bankfull discharge can result.

Entrenched stream types like G and F streams are particularly difficult situations in which to identify bankfull stage. This is unfortunate for stream restoration practitioners since these are usually the streams in need of restoration. Let=s consider a fairly typical evolutionary adjustment pattern. A C channel experiences an increase in its runoff regime causing it to start an accelerated adjustment process. It forms a chute cutoff creating a shortened reach which now has an increased local slope. This section begins to erode headward forming a migrating knick-point, or head cut, as it is known. As it migrates headward it lowers the stream bed creating an incised reach of G channel. As it incises, it abandons its flood plain. As a result larger storms which spread their energy across a flood plain are now confined over the channel without flood relief. The result is more incision. As the knick-point moves headward, excess sediment is discharged downstream of it. The incising reach of channel will not be able to form a flood plain at its now lower level until there is enough material eroded from the channel to make room for flat depositional features to form adjacent to the channel. Identifying bankfull stage in this state is difficult or impossible in a reach such as this.

What does one do? Look upstream and downstream for a reach that can be used to identify bankfull stage; and therefore, estimate bankfull discharge. As the process of adjustment continues the G channel will evolve to an F channel and may begin to develop into a C channel when erosion has removed enough sediment to permit the development of meanders and point bars.

There are some situations where the process of incision progresses in stages. A new flood plain develops and a later change in hydrologic regime causes another process of incision. There was a channel in the Annapolis, Maryland area where three distinct terraces were visible. Age of trees indicate that all of these terraces were formed since the early 1950’s with three distinct AwavesA of land development.

Back to estimating bankfull discharge in these situations, what does one do to be able to differentiate recently abandoned flood plains from current active flood plains. Models such as TR55 or TR20 that predict peak flow from land cover, slopes, a defined storm event, etc., are less than useful for several reasons. They generally overestimate flow from any given storm event used as an input. And there is a problem with picking the right storm event. Should it be a 2-year storm, a 1-year storm, a 12 -year storm or some other event. Gaging station data can be helpful. If regional bankfull relationships have been developed for gaged streams in your region, they can be very useful. If not, you can often select a specific gage with similar land use in the same hydro-physiographic region and field calibrate it in order to apply a proportional area calculation to estimate discharge on your ungaged stream.

(Field calibration of bankfull stage at a gaging station will be covered in a future issue.)