Master of Arts (MA)
Geography & Environmental Studies
Faculty of Arts
Hardy's (1971) technique of interpolation was tested using fluid speed data from Big Otter Creek, a gravel bed stream, in South-Western Ontario. Fourteen cross-sections were selected from a 500 m-long reach of the river. Each cross-section was divided into 10 verticals. Fluid speed was sampled as different positions in these verticals starting at 0.025 m off the channel bed using an electromagnetic current metre. Bed roughness at each cross-section was determined by collecting samples of the bed material and by measuring the long, intermediate and short axes of particles at the 10 verticals. To test the multiquadric technique of interpolation, known fluid speeds from the collected data sets were systematically and randomly removed from the cross-section and plan views. These new, smaller data sets were then used as input for a series of computer programs which generated a multiquadric surface. The interpolated fluid speeds were compared against the know fluid speeds to determine the magnitude of the errors. The results show that the largest error values occurred in the near bed region and in areas with steep velocity gradients. Of the two types of analyses, cross-section and plan view, the plan view had the largest errors. No difference in error magnitude or location was identified between the random and systematic analyses.
Dozois, James Patrick, "A test of the multiquadric technique of interpolation with an application to fluid speeds in a river (Ontario)" (1994). Theses and Dissertations (Comprehensive). 326.