Abstract
The quality of seismic processing and interpretation in the past has always been shackled to the limitations of the acquisition equipment and analysis systems used. The variability of seismic processing between different projects has been due to the geophysicists own personal subjective preferences and interpretations. With the rapid development in computer processing power and technology, this project aims to refine the data processing of a regional data set collected in 1987 with modern methods, in a bid to better the geological understanding of the area and present it in a novel way.
The original 1987 Vibroseis seismic survey involved eight seismic lines encompassing the entire Moutere Depression in Nelson, New Zealand. The target area is an elongated, NNE trending sedimentary basin that is infilled with primarily mid-Miocene to Pleistocene sediments. The age and depositional environment is similar to that of the neighbouring, hydrocarbon bearing Taranaki Basin in the north; it was this similarity that prompted the initial study.
The data processing in this thesis was performed using the Globe ClaritasTM software package with the ultimate outputs being post-stack time migration. The project followed standard seismic practices and used conventional noise reduction methods such as frequency domain (FDFILT) and frequency-offset deconvolution (FXDECON) filters. New frequency-wavenumber and deconvolution filters native to Claritas such as QFK were also tested. Despite a range of filters being trialled, only one type of migration and gain compensator was applied, these were the Finite-Difference Migration and Automatic Gain Control respectively. After processing, the data was then imported into IHS KingdomTM to produce a 3D model of the basin. The six seismic horizons captured during interpretation were defined by their seismic facies. The upper layers were difficult to image due to ground roll and narrow frequency bandwidth sweeps. However, the lower horizons positioned between 500 ms – 2500 ms time depths were interpreted with higher confidence. The strongest reflection occurred with horizon three. The modelling showed a heavily folded and fractured formation with extensive reverse faulting and their corresponding antithetic splays, indicative of a compressive stress environment.
The seismic processing goals of the project was moderately successful. Minor improvements in noise reduction were made using a combination of post- and pre- migration noise filtering. However, due to the poor resolution in the data itself blended with a lack of borehole data in the area, it was difficult to constrain the horizons and interpret the shallow sections accurately. This was a similar issue experienced in the two previous studies of the area. Regardless, my project was able to successfully recreate a 3D interpolated model of the Moutere Depression capturing its significant geological structures such as the Ruby Bay-Moutere and Waimea-Flaxmore faults and as a result, provide a unique perspective of the basin.