4D Chirp-high resolution time-lapse imaging of the marine subsurface

• Main Author: Faggetter, Michael John
• Format: Dissertation
• Language: English
• Published: University of Southampton 2021

Time-lapse seismic imaging refers to the comparison of 3D seismic reflection volumes acquired at different times. This approach has greatly improved our capability to measure and understand dynamic processes in the subsurface. However, there are very few examples using ultra-high frequency(kHz-range) seismic data, especially in marine environments. Exacting requirements for navigation can be prohibitive for acquiring coherent, true-3D volumes and residual errors manifest as noise in time-lapse differences. This is compounded with time-varying amplitudes caused by noise from swell and wash, making it difficult to interpret real subsurface changes. Over coming these challenges opens up a range of applications for monitoring the subsurface at very high resolution. In order to satisfy the requirements for time-lapse imaging, improvements were made to the processing workflow of the 3D Chirp, an ultra-high-frequency sub bottom profiler developed at the University of Southampton. Specifically, methods for post-processing inertially-aided GPS data were incorporated to improve navigation accuracy and stability, minimising the impact of multipath and signal dropouts when surveying in challenging, shallow water environments. This approach provides the ability to acquire repeatable trace data with centimetric-accuracy, absolute positioning, binned onto 25 cm or smaller common-midpoint grids. Following these improvements, two 4D Chirp case studies are presented. In the first, changes to a shallow gas blanket in the Southampton estuary are investigated. Differences are quantified between stacked data acquired at different tidal states. Reflections from the top and bottom of a gas pocket are imaged at low tide, whereas at high tide only the upper reflection is imaged, attributed to hydrostatic pressure-related changes in the free gas. Changes in environmental noise between the volumes were minimized by matching amplitudes at the seabed resulting in a mean repeatability of approximately 40% NRMS, slightly greater than conventional seismic reflection data acquired with towed streamers. The second case study compares two 3D Chirp volumes acquired within a dry dock in Blyth. An artificial sedimentary sequence was disturbed by excavating trenches, burying targets and traversing the site on foot and with machinery. Differences in the acoustic data were used to quantitatively map the anthropogenic disturbance of the dock bed and the shallow subsurface sediments. A comparison of diffractions from an unchanging concrete structure revealed high waveform similarity and amplitude repeatability equal to 5.1% of the stacked volumes. Elsewhere, windowed dock bed RMS amplitudes increased by a mean factor of 3.72, most likely due reduced sediment porosity through compaction. Amplitude differences within the dock were spatially variable and well correlated with centimetric bed-level changes and the location excavated trenches. This relationship was refined by reducing coherent 4D noise: a non-repeatable acquisition footprint resulting primarily from residual navigation errors. This was achieved by applying carefully parameterized trim statics to both 3D volumes, increasing amplitude fidelity and interpretability of the seismic data. These two case studies demonstrate the viability of time-lapse UHF 3D seismic reflection for mapping real, decimetre-scale changes within the shallow marine subsurface, and how a quantitative interpretation of time-varying processes can improve our understanding of complex and heterogenous near-surface sediments and lead to the development of better physical models.