Modeling ocean-bottom seismic rotation rates

Discussion and conclusion

The main conclusion is that rotational sensors can provide us with additional information about the type of waves recorded by ocean-bottom acquisition. The key characteristic for separating between the waves is the ratio between the linear and rotational motions. At the seabed, the P-waves have a mostly linear motion. Scattered S and Scholte waves have a mostly rotational motion.

When designing a new sensor, one of the important questions is its sensitivity. Sensor sensitivity can be parameterized by over-drive level and dynamic range. A sensor that is not sensitive enough will not register weak signal. A sensor that is too sensitive will over-drive too often and will not provide useful information on high amplitude events. Depending on the electronics and the analog instrument impulse response, the time to recover from overdrive may be from a few milliseconds to a second. The dynamic range is of course limited by the analog dynamic range of the sensor and by the number of bits of the digital data (32 bits in modern emerging A/D converters which is usually way above analog dynamic range). In this paper we provide predicted numbers in terms of milliradians/sec of rotation, bars of pressure, and millimeters of displacement.

However, it is important to note that so far our study is based on 2D elastic modeling. In this paper there is circular spreading of body waves and no spreading of surface waves, which is why they do not weaken with offset. The reason we chose the offset $x=1000$$\text m$ at which to analyze the displacements was because the different arrivals were sufficiently separated there, and their different characteristics could thus be displayed. In 3D, there is spherical spreading of body waves and circular spreading of surface waves. The relation between the body-wave rotations and surface-wave rotations may be very different in 3D as a result of the added degree of freedom. We plan to extend this study to 3D before providing sensitivity requirements for dynamic range and over-drive levels.

Modeling ocean-bottom seismic rotation rates