Estimating stream discharge using stage and multi-level acoustic Doppler velocimetry

For temperate region countries with small or moderately sized streams, such as those in Denmark, seasonal weed growth imposes a significant temporal change of the stage-discharge relation. In the past such problems were often avoided by using hydraulic structures, however, firm ecology based restrictions prevent that hydraulic structures are made at the discharge stations presently. As a consequence, the nonlinear drift in weed density and structure adds a significant uncertainty to the hydrograph. Furthermore, the expected increase in extreme discharge situations due to climate changes in the Northern part of Europe may further violate a stable relation between stage and discharge in streams. Extreme high flow situations cause abrupt rise in stage, and consequently weed can be partly uprooted and partly bend down along the bed, thereby changing the conveyance of the stream. In addition, extreme high flow situations can cause the streams to flood the banks. If these hydraulic changes occur in between direct measurements of discharge they are not detected or accounted for in the stage-discharge relation, and the hydrograph can be significantly biased. The objective of this research is to investigate how both seasonal and short duration changes in weed distribution and abrupt changes in stage are recognized in the stream's velocity gradient. It is examined whether the use of multi-level acoustic Doppler velocimetry can provide an improved method for hydrograph estimation with lower uncertainty than traditional stage-discharge methods. In this presentation we shall present results from a study where, at two sites in Denmark, the stream velocity field has been mapped by the use of three Acoustic Doppler Velocity Meter (ADVM) instruments. The ADVM instruments are mounted in three different depths, continuously measuring horizontal average water velocities. Velocity and stage data are selected from one summer and two winter periods, and a method for converting velocity and stage data to discharge will be presented. The estimated discharges are compared with control measurements of discharge obtained from an Acoustic Doppler Current Profiler (ADCP) and with discharge values derived by the Danish Water Authority by traditional stage-discharge analysis. The novelty of this approach is that the velocity gradient is implemented as a marker for hydraulic changes in the hydrograph estimation, without explicit use of cross sectional area and cross sectional average velocity. The initial results show good agreement between the ADCP discharge measurements and the discharge values predicted from the stage-velocity method and suggest superiority of the stage-velocity method during abrupt changes in stage.