Figure 1. Random samples of velocity and depth in stream reaches.
WATER RESOURCES RESEARCH, VOL. 38, NO. 1, 10.1029/2000WR000014, 2002
|
|
[6] This study uses 149 reach surveys for streams in Australia, New Zealand, South Africa, and Europe (see Tables 1 and 2). Each reach survey provides a set of velocity-depth pairs that represent a sample of hydraulic conditions throughout the reach at the time of the survey (Figure 1). These surveys were mostly to provide input data for habitat modeling, and, although these data have been supplied by a variety of agencies, the survey methods used to collect the data are similar. Samples generally consist of velocity and depth measurements taken across several cross sections with velocity measured at 0.4 times the depth above the bed (0.4Z) or, in a few cases, at 0.2Z and 0.8Z. Depending on the number of velocities available, the depth-averaged velocity is estimated using
or
In many cases the location of measurement points across the channel is not random; rather it is to provide a good description of lateral variations in velocity and depth. As a result, fewer measurements are taken where conditions are relatively uniform. To account for this nonrandom sampling strategy, it is common practice to weight measurements, using the width of a segment of the cross section occupied by the measurement, when calculating reach statistics from the sample values [Bovee, 1994]. For this study, segment width is estimated as half the distance between adjacent measurement points. Discharge for the 149 surveys ranges from 0.01 m3/s to 100 m3/s, and sample size varies between 20 and 500 depth-velocity pairs. These surveys encompass a broad range of stream types in upland and lowland streams and reaches subject to a variety of human disturbances including flow regulation.
Citation: A stochastic model of hydraulic variations within stream channels, Water Resour. Res., 38(1), 10.1029/2000WR000014, 2002.