An assessment of variants in the professional judgement of geomorphologically based channel types

Abstract

River classification is a useful tool for researchers and managers wishing to organise, to simplify and to understand the forms and processes within freshwater systems. Many classifications require surveyors to classify reaches into specific channel types in a field environment. Channel types should be identifiable on the basis of a field surveyor's judgement of channel characteristics and landscape settings; these include channel planform, valley confinement, dominant bed material and/or instream geomorphic features (e.g. gravel bars). An accurate classification of reaches into the correct channel type is important to ensure consistency in management strategies and to assess the impact of engineering activities on the physical and ecological status of rivers. In this article, we examine the variation in professional judgement of geomorphologically based channel types by scientists with different disciplinary backgrounds and varying levels of involvement in classification systems using a photo-questionnaire. Results indicate that there can be a large level of discrepancy in typing rivers; the choice of the modal channel type for each reach varied between 25.9% and 75.1% of the respondent selections. There were also differences in the level of agreement between earth scientists (with hydrogeomorphological or geological training), ecological scientists (with freshwater biology training) and practitioners involved in river conservation and management. A high level of experience in classification systems translates to a lower number of channel types being chosen per reach. In response to these results, the use of a photographic approach to typing needs to be fully tested and users fully trained before operational use. Furthermore, we advocate that designers of geomorphic typologies should aim to have a representative and workable number of classes within a typology with an emphasis for rationalisation of classes rather than expansion of numbers.