• Confluence;
  • Rivers;
  • Discordance;
  • Glaciers;
  • Hydraulic geometry


Geomorphologists since Playfair have apparently accepted that fluvial confluences will be morphologically accordant in distinction to the hanging junctions of glaciated troughs, a view reinforced by Gilbert's ‘Law of Declivities’. The issue has been largely ignored despite field instances of discordant fluvial junctions and widespread engineering practice which favours discordant confluences in non-erodible materials.

Six categories of fluvial junctions are defined: accordant, hanging (discordant), barred, impeded, disrupted and disturbed. The study focuses principally on the first two classes and defines discordance purely morphologically in terms of relative depth of incision of the confluent thalwegs. Likely controls of accordance/discordance are identified as flow magnitudes, synchroneity of flow régimes, channel boundary materials and debris calibre.

Field investigations in Malawi, France, West Malaysia and New Zealand in small (< 100 km2) basins on a variety of rock types are described. Barred junctions are relatively rare, but complex in character. A high proportion of confluences are distinctly discordant and there is a general continuum from accordant cases. Tectonic control is improbable and the principal cause identified is the degree of difference in channel-forming discharges, estimated by stream magnitude and channel area. Some evidence for lithologic control is also apparent. It appears that many confluences behave as if the tributary junction is non-erodible and not in the fashion predicted by the Law of Declivities. It is argued that the observed relation is a necessary result of the general down-stream changes in hydraulic geometry.

Limited comparison with glacial trough confluences reveals a difference in degree, but not in nature, although the relationship between discordance and channel size is affected by cross-section geometry. A 10-fold difference in the cross-sectional areas of channels meeting at a junction produces, on average, a 2·4-fold difference in mean channel depths for fluvial cases and a 4-fold difference for glacial troughs.

The role of varying flow regimes seems to require attention. Nevertheless, it is maintained that ‘Playfair's Law’ is a misleading description of the morphology of fluvial junctions, which behaves fundamentally in the same fashion as glacial troughs.