Quantifying 'effective' floods using large-sample geomorphology

Geomorphologists have long debated the relative importance of disturbance magnitude, duration and frequency in shaping landscapes; for channel change during single floods, it is thought that flood duration, rather than magnitude, matters most. However, studies of flood-induced channel change have often drawn upon small datasets. By using satellite data to track channel adjustment during floods, we can now query these classic hypotheses with large datasets, and we do so here by combining 7 years of Sentinel-2 images with daily flow data from laterally active rivers. Using Earth Engine, we apply automated algorithms to map river planforms and detect their lateral shifting, and we generate a large dataset to quantify channel change during ~1000 flood events in gauged rivers across New Zealand and the Americas. We draw upon this dataset to evaluate how characteristics of the flood hydrograph (including magnitude, duration, and integrated sediment transport) correlate with the degree of geomorphic change observed. Finally, we examine the potential of predictive models for geomorphic change during floods, and consider the variables that moderate this relation between flood character and geomorphic change.