The energy responsible for the features in the raw data, Figures - , are likely related as they share a locus at the production facilities and are both characterized by water velocity. I suspect that the noise-train in Figure is a collection of aliased repeating events similar that found in Figure . Also, the gathers generated with data showing the diffuse noise-train, Figure , have almost identical kinematics to the raw data showing a crisp water-velocity event, Figure . It is possible that some activity on the platforms could sometimes be impulsive in nature or more drawn out. In either case, the correlated gathers do not show convincing evidence that the energy penetrates the subsurface after the direct arrival is recorded. Forward modeling the direct arrival to the receiver stations shows that water velocity events are aliased close to the source with this array. This analysis of the raw data also shows that the predominance of observable energy shares a single source location.
Correlating the raw data to produce synthesized shot gathers was always performed on 12 second records. These individual results were summed from 1-2500 times to generate some approximation to active seismic shot gathers in accordance with the theory of interferometric imaging. The gathers produced with a single 12 second record, Figure , or 15 hours of data correlated in 12 second sections, Figure , show an identical strong event with velocity 1450 m/s arriving from the Ardmore field 40 km to WSW. Less obvious events are also present in the gathers with locus at the Valhall production facilities in the center of the array. These events are also characterized with water velocity.
It seems counter-factual to attribute the source of this energy to a distant production operation when major production, drilling, and injection activities within the array have not generated events that match the power and bandwidth of this most obvious arrival. The British Department of Trade and Industry describes the Ardmore development. It differs from the Valhall operation in several important ways. The Ardmore facility has no permanent structures, nor is it connected to an export pipeline. Instead, two vessels are permanently moored over the field: One contains production facilities, the other is used to store liquids for periodic transport to shore. This operation has no way to handle produced gas and must therefore flare gaseous production to the atmosphere. I believe this is the most significant difference between the two operations, and that the gas flare is the most likely sonic source for the energy recorded at the array.
With only two main energy sources developing the events in the correlated gathers, the result does not satisfy the equation which relates passive recordings to active data acquisition. Correlating events from a single source simply moves the minimum travel-time captured by the array to zero. Constructive and destructive interference from many events are required to generate events with the kinematics of an active survey. Sorting the gathers to midpoint-offset coordinates yields completely uninterpretable CMP gathers. Similarly, further processing which assumes impulsive sources located at the trace used as source function to synthesize a gather (migration) is therefore inappropriate. However, having identified the locus of the energy in the correlated gathers, the equation describing the kinematics of the event can be used as a mute function or areal source function for further processing.