1 Comparison of Automated Grain Sizing of Gravel Beds Using Digital Images to Standard Grid and Random-walk Pebble Counts

An automated grain sizing (AGS) procedure which measures grain size distributions from digital images of river beds was tested to see if it could be used as an acceptable alternative to standardized random-walk pebble counts for measuring grain size information in gravel-bed rivers. Data for the comparison was collected on six different field sites using three different grain size measurement methods. The image processing methodology was then applied to the experimental field data to examine the variability and accuracy of the method compared to the grid and random-walk pebble counts. Results showed that: 1) the AGS method produced less at-site variability from sample to sample when compared to the grid and random-walk pebble counts for all six rivers; 2) the AGS method produced grain size distributions that were more accurate in relationship to the grid count statistics than the random-walk method for grain size statics d5-d50, whereas the random-walk method produced statistics that were closer to the grid count method for grain sizes d84-d95; and 3) the variation between the AGS and the grid method was less than 0.25ψ for data from all six sites