Characterizing the relationship between river branch number and water
discharge using distributions of channel belt properties
Abstract
Distinguishing single-thread and braided channel patterns and
reconstructing paleo-water discharge from the sedimentary rock record
has proven to be difficult. This is because only remnants of the river
channels are preserved, often accounting for a small fraction of the
overall stratigraphy. Instead, channel belt deposits—the amalgamation
of deposits from many individual channels—are observed more often.
Identifying channel patterns from geologic records is important for
testing a range of hypotheses, including whether single-thread rivers
were rare prior to the arrival of land plants in the Silurian. Here, we
develop new quantitative metrics using distributions of channel geometry
and channel belt properties with the ultimate goal of distinguishing
channel types in the rock record. Metrics measured include width,
sinuosity, radius of curvature, wavelength, and amplitude, and are
measured from modern fluvial systems in the US via remote sensing
techniques. An entropy-based braided index is used to quantify river
branch number, which is directly related to a conventional braided
index. Preliminary finding suggest that channel belt width and
wavelength are well correlated with river branch number and water
discharge via power-law relations, respectively. Specifically, mean
river branch number is inversely related to channel belt wavelength,
normalized by mean total channel width. In addition, mean channel belt
width is directly related to water discharge at 2% exceedance flow
event. This could imply that extreme flow events may contribute
significantly to channel belt formation, more so than typical bankfull
flow events (5–25% exceedance flows) based on threshold channel
theory. Analyses presented here will aid improved interpretation of
sedimentary deposits on Earth and other planets to infer past surface
conditions.