24th May 2017

When the going gets rough…
I recently gave a sneak preview of my EAGE conference paper “comparison of methods for rough sea-surface estimation” to Shearwater colleagues as part of our regular series of in-house ‘Lunch and Learn’ seminars. If the pleasure of learning about the effect of rough seas on seismic data was not sufficient inducement, there was also a selection of free pizza on offer. 

So, what was the seminar about? You really should attend the Paris conference next month for all the detail (broadband session 14 June) but just in case you can’t then, in short, the research is about how we can do a better job handling seismic data acquired in rough sea conditions. 

Our vessels tow long streamers of hydrophones behind them, which are designed to record an extended series of echoes coming from beneath the Earth’s surface. This is the seismic data that we want. However, the hydrophones also record a slightly delayed second ‘copy’ of the seismic data that has been reflected at the sea-surface. This is called the ghost and we aim to remove it from the data, as it is frankly confusing for any geologist who is expecting to see only one Earth. For the purposes of calculating the ghost response, it is simplest to assume a mirror-like reflection. This is ok if the sea is calm, but is not a good assumption when the sea is rough. In practice, by ‘rough’ we are talking about wave heights exceeding perhaps as little as half a metre. The waves passing above the streamer perturb the arrival time of the sea-surface ghost reflection and this makes de-ghosting tricky. It can lead to unwanted signal processing artefacts if the timing variations are not known. 

We have recently developed an algorithm that handles variable sea-surface and streamer depth profiles explicitly, so making a reliable estimate of the sea-surface profile is the remaining piece in the puzzle. That is the focus of this most recent research, co-authored with Sergio Grion, where we have compared two different and complementary approaches for deducing the sea-surface profile from the recorded data – one via an interference method and one using the linear theory of wind-driven waves. The refinements to processing that this extra information brings will enable us to maintain a high quality product, even when the weather is not on our side. 

Rob Telling, Senior Research Geophysicist