Automatic detection of river capture based on planform pattern and χ-plot of the stream network

Abstract

River capture is a surficial process that can lead to drainage reorganization and have significant impacts on sediment dispersal and biotic evolution. Discovery and study of river capture events are helpful in revealing the history of drainage basins, however, the present-day identification of river capture mainly depends on field and map observations by geomorphologists, who lack an automatic method. In the case of a large-scale drainage system with a complicated stream network, a field investigation is too time-consuming and costly. This study aims to develop a novel method for automatic river capture detection based on planform morphology and χ-plots of the stream network. The whole method can be described as a workflow including three steps: (1) searching candidate regions where river capture may have occurred; (2) iterating over each candidate region and roughly detecting the captor, captured, reversal, and beheaded rivers using the planform pattern of the stream network; (3) verifying the river capture with a χ-elevation plot in the candidate region found by rough detection. The description of the method is followed by two case studies from China and Spain, demonstrating the ability of the method to identify the location where the river capture might occur. We also discuss the parameters that must be optimized before extracting the river capture efficiently from the stream network.