In the 1960s, S.A. Schumm proposed a set of conceptual relations to predict river channel adjustment. These relations take into consideration runoff and sediment load variables only; when channel boundary materials are changed considerably during adjustment, it is likely that these relations would not work well. After a discussion about some limitations of Schumm's conceptual model, this study introduces the channel boundary material as an independent variable, and thereby establishes a conceptual model.

This model can be used to predict the tendency of downstream and upstream channel adjustment after reservoir construction, and particularly to explain the complex behavior of channel adjustment at a relatively long time scale. Although this model is qualitative, it proposes a framework or a line of thinking for further physically based simulation modeling. This conceptual model has been tested by data from the Hanjiang River and the laboratory model river, and also applied in the navigation channel training practice.

The suggested conceptual model is flexible in application, and may be used in rivers with various physico-geographical settings. For example, when the bed material is fine and selective scour is less marked or duration of the clear water scour is short, and the direction of channel adjustment can be predicted by the first stage of this model; in fact this is just Schumm's model. f the bed material coarsening is pronounced, the re-shaping of point-bars is also appreciable, then the channel adjustment may rapidly enter the second and the third stages. The complex behavior of channel adjustment can be explained satisfactorily.