Forest Investment Account (FIA) - Forest Science Program
FIA Project Y071074

    Hydrologic Indicators for Watershed Sensitivity to Peak Flow Changes in Small Watersheds
Project lead: Hassan, Marwan
Contributing Authors: Hassan, Marwan A.; Brayshaw, Drew
Imprint: Abbotsford BC :, 2007
Subject: Forest Investment Account (FIA), Hydrology, British Columbia
Series: Forest Investment Account (FIA) - Forest Science Program
One of the major challenges in mitigating or predicting the impacts of disturbance on hydrologic systems is to predict changes in both hydrology and in sediment transport in channels. Typically, only one of these components is considered in detail and the effect on the other component is inferred. In particular, much of the research on the effects of forest harvesting on hydrology has focused on evaluating changes in streamflow without considering the changes in the morphology of streams in which the flow occurs. For the purposes of forest management, linking hydrology and sediment transport is of crucial importance in order to improve management practices and to move from statistical and deterministic modeling to management outcomes with observable results. This is especially true of the current era in British Columbia, where widespread changes in forest cover resulting from mountain pine beetle infestations and associated salvage logging are affecting the hydrology of streams over scales ranging from small watersheds up to large regions of the province at once. The shape and dimensions of stream channels are formed not only by the water flowing from upstream and from hillslopes, but also by the sediment flux from upstream and from adjacent hillslopes and streambanks. This is especially the case for smaller streams that are directly connected to the adjacent hillslopes. Therefore, the state of the channel, its form and dimensions can be used as an index for the streamflow, sediment flux, and antecedent conditions which have shaped the channel. Bankfull discharge and effective discharge are two characteristic parameters of the channel, which can be used to evaluate the sensitivity of watersheds to disturbance (Nolan et al, 1987; Goodwin et al, 1998). The bankfull discharge is commonly defined as that discharge in which an alluvial stream is full to the top of the bank without overtopping it (Williams, 1978), although numerous alternative methods of definition exist based on morphologic, biologic, and statistical analysis methods. In formerly glaciated terrain, such as British Columbia, the top of the bank may have been formed during a flow regime which no longer exists. Therefore the bankfull discharge, although relatively easily measured, must be interpreted with caution. In general, the bankfull discharge is assumed to be a common event. Wolman and Miller (1960) assumed the bankfull and effective discharge to be equal to the mean annual flood. However, they warned that this would not be the case for smaller upland streams. Dunne and Leopold (1978) considered the return period for bankfull discharge to be approximately 1.5 years. However, Williams (1978) examined fifty studies and found that the reported return period for bankfull discharge ranged from a few months to 200 years. Castro and Jackson (2001) reported that for large watersheds (>100 km˛) in the Pacific Northwest, the mean return period for bankfull discharge varied between 1.2 and 1.5 years, depending on climate, but that some watersheds varied significantly from the mean. In British Columbia, forest management guidelines (MoF 1999, 2001) assume that bankfull discharge has a return period of 2 years for all streams regardless of size. Researchers have attempted to relate the bankfull discharge to channel and basin characteristics in order to predict the bankfull discharge. Kilpatrick and Barnes (1964) found that as channel slope increased, so did the return period of bankfull discharge. Williams (1978) also found that as channel gradient steepened, the return period for bankfull discharge increased. Andrews (1984) found that the return periods of bankfull and effective discharge increased as watershed area decreased. Castro and Jackson (2001) found that within nested catchments, the return period generally increased when moving from a downstream to an upstream gauge. However, in low-gradient headwater streams, Dodov and Foufoula-Georgiou (2005) found that bankfull discharge was more frequent as watershed area decreased. The wide range of results highlighted above makes transfer of results from one region to another problematic; none of the published work may be applicable to British Columbia. Another parameter used to characterize stream channels, which is potentially more useful for linking changes in hydrology to changes in sediment transport, is the effective discharge. The effective discharge (or dominant discharge) is defined as that discharge which transports the most sediment and thus is most effective in forming the channel (Wolman and Miller, 1960). In some cases, the two parameters are equivalent; the bankfull discharge will also be the effective discharge (Wolman and Miller, 1960; Dunne and Leopold, 1978; Andrews, 1980). In other cases, the effective discharge has been assumed to be equal to the bankfull discharge without supporting evidence. Since effective discharge is a parameter related to sediment transport within a given reach, it is intuitive that the frequency of effective discharge will decrease as channel sediment size increases. Emmett and Wolman (2001) showed that for sand-bed rivers, the frequency of effective discharge generally equaled that of bankfull discharge, while for coarser gravel-bed streams, the return period of effective discharge was about twice that for bankfull discharge. Costa and O’Connor (1995) demonstrated that the shape of the hydrograph and duration of overbank flows affect the effective discharge. For small upland watersheds in the Oregon Coast Range, Grant et al. (1990) found that the return period of channel-forming events was 50 years or greater. Church (2002) discussed reasons why the effective discharge should be an infrequent event in smaller watersheds. Nolan et al (1987) found that in aggrading and degrading streams in the California Coast Range, the effective discharge and bankfull discharge could have return periods that differed by an order of magnitude or more. Goodwin et al (1998) suggested the use of the bankfull and effective discharges as index parameters to guide stream restoration works, although they assumed that in a stable stream the effective discharge would equal the bankfull discharge. In formerly glaciated terrain such as British Columbia, where there may be significant lag material within the channel, the return period of effective discharge may be much greater than the return period of bankfull discharge. There is a lack of strong evidence to suggest that bankfull discharge and effective discharge will be the same for small headwater streams in British Columbia. Currently, there is no good method for predicting the return period of bankfull or effective discharge for most small watersheds in British Columbia. The research cited above, which represents the scope of current knowledge but is not intended to be exhaustive, may therefore not be applicable to British Columbia, even considering research from closely adjacent areas such as Washington state. Furthermore, there are regional differences within BC such that results developed for a single region of BC might not be applicable to other regions. Exploring regional relations of the effective and bankfull discharges is therefore of great practical importance for BC, in order to predict the regional response patterns of streams to natural or non-natural disturbances. The effects of changes in forest cover on forest hydrology, whether resulting from forest management practices, forest fires, or mountain pine beetle infestations, can affect watershed hydrology, particularly by increasing the frequency and magnitude of peak flows. While the details remain subject to debate, and ongoing research continues, there is a general consensus (Bechsta et al, 2000) that the effects of changes in forest cover on small peak flows (return period less than 5 years) is greater than for larger peak flows. Therefore, in order to understand the potential effects of changes in forest cover on stream channels, particularly in small watersheds, it is necessary to understand the characteristic return periods of bankfull and effective discharges in those watersheds. If these channel-forming flows are frequent events with low return periods, then the frequencies with which bankfull discharge and effective discharge occur can be expected to change as a result of widespread beetle kill, or forest fire, or extensive forest harvesting. Such a change in frequency could be expected to produce measurable effects on streambank stability, stream channel morphology, water quality, and riparian and aquatic ecosystems. On the other hand, if bankfull and effective discharge are relatively infrequent in small streams and correspond to relatively large, rare floods, then small streams might be considered to have relatively lower vulnerability to increasing peak flows. In this case it is possible that significant changes in forest cover, while increasing peak flows and water yield, might not result in significant changes in sediment transport or stream channel pattern. The goals for this project are therefore to characterize the range of return periods for bankfull discharge and effective discharge for small headwater streams in British Columbia, to evaluate the scaling relations for bankfull and effective discharge in small streams, and to assess the effectiveness of bankfull and effective discharge as indicators of watershed sensitivity to changes in peak flow.
Related projects:  FSP_Y082074FSP_Y093074
Contact: Hassan, Marwan, (604) 822-5894,


Executive Summary (0.1Mb)

To view PDF documents you need Adobe Acrobat Reader, available free from the Adobe Web Site.

Updated August 16, 2010 

Search for other  FIA reports or other Ministry of Forests and Range publications.

Please direct questions or comments regarding publications to