Measurement of sediment loads during flash flood events: 14 years of results from a six stream monitoring network on the southern Colorado Plateau

Abstract:

In in arid and semi-arid environments, short-duration, high-intensity rainfall events—flash floods—are the primary driver of sediment transport in ephemeral streams. The spatial and temporal variability of these rainfall events results in episodic and irregular stream flow and resultant sediment transport. As a result of limited-flow durations, measuring discharge and collecting suspended-sediment samples on ephemeral streams in arid regions is difficult and time-consuming. Because of these limitations, few sediment-monitoring programs on ephemeral streams have been developed; some examples of sediment-monitoring gages and gaging networks constructed on arid ephemeral streams include Walnut Gulch, United States, Nahal Yael, Israel, and the Luni River Basin, India. The difficulty in making measurements of discharge and suspended-sediment concentration on arid ephemeral streams has led many researchers to use methods such as regional sediment-yield equations, sediment-rating curves, and peak discharge to total-sediment load relations. These methods can provide a cost-effective estimation of sediment yield from ungaged tributaries. However, these approaches are limited by, among other factors, time averaging, hysteresis, and differences in local and regional geology, rainfall, and vegetation.

A monitoring network was established in 2000 on six ephemeral tributaries of the Colorado River in lower Glen and upper Marble canyons. Results from this monitoring network show that annual suspended-sediment loads for individual streams can vary by 5 orders of magnitude while the annual suspended-sediment load for the entire network may vary annually by 2 orders of magnitude, suspended-sediment loads during an individual flood event do not typically correlate with discharge, and local geology has a strong control on the sediment yield of a drainage basin. Comparing our results to previous estimates of sediment load from these drainages found that previous, indirect, methods that used no actual measurements of sediment transport greatly overestimated the annual suspended-sediment loads of the streams.