A REMOTE SENSING METHOD FOR DETERMINING BANKFULL DISCHARGE

In principle, bankfull discharge (i.e., the discharge that just fills a channel’s banks) is easily observed in the field by identifying the geomorphic banks as part of a field survey, monitoring these banks continuously until they are inundated, and then recording the discharge at that time. In practice, this procedure requires a significant amount of field effort and can be subjective. Consequently, bankfull discharge is often estimated using frequency analysis of annual peak flow data, with the 1.5 year discharge used as a proxy. Alternatively, breaks in the relationship between stage and discharge at a stream gage may be used to identify bankfull stage. However, this method assumes that the gage is installed in a geomorphically representative location, which is not always the case. Remotely-sensed aerial photography can be used to estimate bankfull discharge in a way that is completely independent from traditional approaches. The method requires three steps. First, wetted channel planform area for a reach several bends long is measured across a range of discharges from a series of aerial photographs. Second, the bankfull channel area (i.e., the total area between vegetation lines) is determined from the photographs. Finally, a simple power function is fit between the average wetted width (wetted area/reach length) and discharge as observed on the date of each respective photograph at a nearby gage. The value of the power function at the average bankfull width represents the bankfull discharge. The method was applied to the 9.8 km reach of the Le Sueur River, Minnesota, between the confluence of the Blue Earth and Maple Rivers. The bankfull discharge estimate is required as input into a larger sediment routing model being developed for the system. Six photographs were available for the period between 2003 and 2008. The average wetted width for these photographs ranged from 62% to 99% of the bankfull value, with all but two points below 81%. The bankfull estimate of 98.6 m³/s obtained using all the data is within 2% of the estimate made without the two largest discharges (the largest is almost a direct measurement of bankfull discharge). The result is consistent with ground-based photographs (slightly higher than bankfull at 113 m³/s) and with the traditional 1.5 year discharge estimate (~96.3 m³/s).