SPECTRAL ANALYSIS OF RIVER SINUOSITY FOR IDENTIFYING INTRAPLATE CRUSTAL DEFORMATION

The South Anna River in central Virginia shows anomalous meandering co-located with the epicentral region of the 2011 M_w5.8 Mineral earthquake. Similar hydraulic geometry changes of alluvial channels, in response to changes in tectonically-induced valley gradient, are well documented in both field and modeling studies. However, characterizing the response of bedrock or mixed alluvial-bedrock channels, and the need for robust, reproducible, and objective ways of quantifying meander behavior, are nontrivial. To address these challenges, we explore numerical routines to objectively measure the frequency and amplitude of river channel meandering, building off previous analyses of the South Anna River. We select a suite of rivers (South Anna River, VA; Ashley River, SC; Boeuf River, LA; Ouachita River, LA) coincidental with known or suspected regions of intraplate crustal deformation to test our numeric technique. We extract coordinate points every 5m along the channels from a 30 m-resolution digital elevation model digitized in ArcGIS. We first smooth the points with designated running average window lengths (2km, 4km, 8km) to remove the large-scale river course. We then resample to an interval of 25m, with the amplitude being the deviation of the river channel from the smoothed course. The resampled channels are analyzed in the frequency domain after a Fourier transform, and the output is an evolutionary spectrogram (a plot of meander frequency and amplitude vs. distance along channel) that can be georectified to the original planimetric river channel. Correlation between large-amplitude signatures on the evolutionary spectrogram and regions of uplift or subsidence caused by crustal deformation suggests that there are tectonically-driven gradient changes in the valleys of these rivers. However, other channel reaches of high spectral power do not correlate with a known tectonic source. These latter anomalous reaches deserve further investigation to understand the causative factors that change meander amplitude and wavelength, including changes in rock type and far-field base level fall. We freely provide the numerical code (written in GNU Octave) to encourage the continued evaluation of river meandering analysis method, and further explore the factors influencing meandering in actively deforming regions.