We examine two different non-gimballed 4-component ocean-bottom-cable datasets that were acquired using identical GeoRes-SubSea cable systems. In the Seneca Lake survey, the cable was laid out untrenched. In the Valhall LoFS (Life of Field Seismic) baseline survey, the cable was trenched to 1m depth. As expected, the untrenched dataset shows a pronounced imprint of the cable direction on the vector fidelity of the direct arrival, while the trenched-cable dataset shows excellent vector fidelity throughout.
A few receivers in the untrenched dataset do show notably better vector fidelity, however. These appear to be exactly the receivers located at points where the cable was laid out strongly curved. We conclude that with a suitable cable layour, it may be possible to damp down the undesired rocking mode of the cable without trenching. However, even the best receivers in the untrenched dataset are still inferior to typical receivers in the trenched dataset.