Knock Knock, Who's There? Identifying the Source of Signals in the First DAS Deployment in a Fluvial Setting

Important fluvial parameters such as turbulence and sediment transport can be difficult to measure directly and continuously. Over the decades, many studies have developed traditional seismic methods to try to monitor changes in these parameters. Distributed acoustic sensing (DAS) is a novel seismic sensing technique that utilizes fiber optic cable to record data comparable to what would be achieved by a large seismic array of single channel geophones or accelerometers. Many studies are beginning to surface about DAS's utility and possible applications across the geosciences. However, the application of DAS in fluvial systems is only beginning to be explored, and we analyze data from what we believe to be the first DAS deployment in a flowing stream: a recent deployment in Clear Creek in Golden, CO. Whether using DAS or more conventional seismic instrumentation, a critical challenge in analyzing and interpreting fluvial parameters from seismic data is the overlap of signals produced by numerous varying sources and processes. In our case, this is compounded by the fact that the fiber optic cable was not fixed to the creek bed, meaning that motion of the cable itself is another source of signal. An analysis of the Clear Creek DAS data reveals different types of signals with distinct spatial and spectral patterns. One especially interesting signal is seen as a series of impulsive arrivals and can be heard when the data is played as audio as a series of "knocks" reminiscent of pebbles or cobbles rocking in place. We investigate these signals and find that A) they consistently originate at the point where the cable is closest to the bed, B) The signal propagates at a speed of ~2,000 m/s, and C) The signal undergoes total reflection at a point along the cable. The high propagation velocity rules out the possibility of the signal being sound propagating through the water and, along with the observation of total reflection, suggests the signal is propagating through the cable itself. The fact that the signal consistently originates where the cable is closest to the bed suggest the cable is hitting the bed, causing the impulsive signal that then travels through the cable. At a point upstream where the cable seems to have draped over a boulder, the signal reflects. Our study brings us one step closer to understanding what a DAS fiber might record when deployed in a flowing stream and helps plan future deployments.