Identifying pressurized flows under river-ice using seismology: insights from a flume experiment

River-ice affects hydraulics and sediment transport that may in turn influence channel morphology. However, scientific understanding of sub-ice flows is limited by the difficulty of accessing the ice-covered channel bed and banks. During periods of stable ice cover, hydraulic studies usually assume that the stable ice cover is free-floating and can therefore move vertically to accommodate changes in river discharge. However, ice cover is often fixed in place, attached to the channel banks. In this case, increasing discharge is forced under the ice cover causing pressurized flows typified by higher flow velocities and sediment transport. The identification and study of pressurized flows is difficult due to the challenges of measuring flows in ice-covered rivers during high discharges; in particular since common methods of drilling holes to measure velocities will disrupt any potential pressurization.

We aim to determine if environmental seismology can be used to identify pressurized flows in rivers and to interpret the characteristics of seismic signals to inform knowledge of hydraulic processes during pressurized flow events. Thus, we set up a flume experiment to compare the hydraulic seismic signature of free-surface flow with pressurized flow under fixed ice-covered conditions. Using a 7m-long transparent 10 x 10 cm PVC tube and fixing roughness elements onto the riverbed (sand and gravel), we test three configurations varying the discharge and the distance between the bed and the bottom of the ice cover (simulated by the upper surface of the inside of the tube). The slope is 0.3 % to represent prototype low-slope subarctic river channels. Two PE6/B three-component 4.5 Hz geophones record millisecond resolution seismic data: one is installed on top of the water-filled flume, and the other on an empty 1m-long section of the same type of PVC tube placed next to the flume, to record background noise. We can pressurize the water-filled flume by increasing the discharge for a given treatment, and record discharge and video data to identify and describe pressurization events.

Comparing seismic and discharge data confirms that we can identify hydraulic signals in the seismic record. We observe a scaling relationship between discharge data and seismic power, and are investigating its coherence with existing theoretical models and its dependency on apparent bed roughness. We expect pressurized flows to appear as high-energy signals due to increased water velocity, with a decrease in background noise due to complete contact between the water and the pipe.

These results can help resolve a long-term aim of identifying the occurrence of sub-ice pressurized flows from seismic field data. Such understanding has implications for using seismic signals to calculate stage in ice-covered rivers or subglacial channels and calculating ice-related bedload transport. These techniques provide unparalleled opportunities for non-intrusive and continuous measurements of hydraulic processes under ice.