Source Assisted Marine Refraction Microtremor (ReMi) for Marine Material Strength Assessments – New Ireland Province, Papua New Guinea

Abstract

Refraction Microtremor (ReMi) is a relatively new technique which utilises ambient noise generated from urban infrastructure in addition to the natural seismic energy to generate shear wave models of the subsurface. This paper presents the geological setting, field acquisition parameters, problems and limitations of data collection and results of a survey conducted in a shallow marine harbour in Papua New Guinea. The primary objective was to undertake an assessment of the viability of Marine ReMi to obtain subsurface parameters to assist in an overarching geotechnical study taking place on site. The subsurface parameters that the investigation was aiming to define included shear strength and stratigraphy. The environment in which this trial was undertaken was a shallow marine environment containing paleochannels, coral, marine sediments and landslide material. The field setup used was designed to target shallow features and establish their shear wave velocities. The setup consisted of using 10Hz Hydrophones at 2m separation, with a record length of 20 seconds stacked 8 times. During collection it was found that relying on ambient noise alone was insufficient to gain the required data resolution. To provide additional energy a controlled source in the form of a bolt airgun was utilised with shots being taken at arbitrary locations off the line and up to three times each stack. Inversion of the data was undertaken as per standard passive seismic inversion. The results of the investigation show that through careful processing and analysis accurate models of S-wave velocities beneath the ocean floor can be successfully achieved using ReMi. We were able to cross correlate the results with traditional marine geophysical techniques, land based geophysics and physical observations to gain a better understanding of the accuracy of the modelled Marine ReMi data compared with other data streams. This correlation shows that the results are comparable in both structures and velocities to the other data streams.