Non-uniformity of local environmental conditions leads to differences in the axial orientation, length, and degree of development of individual meander bends, so that rivers vary in meander irregularity as well as in sinuosity and dominant wavelength. Wavelength estimation is therefore necessarily statistical; various methods are compared for nineteen British river reaches digitised as direction series. For a given reach, mean bend length is strongly dependent on the sampling interval over which each direction measurement is made. Direction spectra are not significantly polymodal, in contrast to Speight's findings, but dominant wavelengths are high or infinite because of valley bends. Better wavelength estimates are obtained from direction-change spectra, whose single peaks are not systematically affected by the choice of sampling interval; lack of resolution can be partly overcome by averaging estimates for different sampling intervals. Direction and direction-change autocorrelograms are generally of damped oscillatory form, and wavelengths estimated from direction-change autocorrelations correlate closely with spectral estimates. Both wavelengths are nearly linearly related to channel width and the square root of 1% duration discharge, irrespective of flow variability, width/depth ratio, or slope. /Author/

  • Corporate Authors:

    North-Holland Publishing Company

    P.O. Box 211
    1000 AE Amsterdam,   Netherlands 
  • Authors:
    • Ferguson, R I
  • Publication Date: 1975-8

Media Info

  • Features: Figures; References; Tables;
  • Pagination: p. 315-333
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00125018
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Oct 18 1975 12:00AM