Abstract
Natural gas hydrates act as an efficient natural sequester of large amounts of carbon, and it is estimated ~15% of Earth's total mobile carbon could be stored in gas hydrate provinces. They form complex systems in the shallow sediments of deep-sea regions, where there is sufficient supply of natural gas in the sediments, and the conditions of low temperatures and high pressure at which gas hydrates are stable are met. Most gas hydrate provinces are found on continental margins. The southern Hikurangi subduction margin, offshore Wairarapa (New Zealand), contains a large gas hydrate province, inferred by the presence of widespread bottom simulating reflections (BSR) in shallow sediments across the margin. Moreover, locally intense fluid seepage at the seafloor associated with methane hydrate has also been observed and documented. Understanding the complexity of hydrate systems is valuable for a range of scientific issues related to climate change and ocean chemistry, geological hazards, deep-sea ecology, and energy supply. A quantitative approach to the characterization of gas hydrate systems in the region is an essential step towards the estimation of the local carbon budget, especially in terms of the total volume of gas hydrate in the region and estimation of gas fluxes through the seafloor at cold seep locations.
In this thesis, I combine a range of geophysical data and theoretical models to identify, map and quantitatively characterise gas hydrate deposits and cold seeps on the southern Hikurangi Margin. Multi-channel seismic (MCS) data and methods form a large part of the basis of the studies presented in this thesis. Two datasets were used: long-offset, lower resolution industry data APB13 (R/V Duke, 2013) and higher resolution, short streamer data TAN1808 (M/V Tangaroa, 2018), acquired as a densely spaced grid over five target sites. The synthesis of these datasets provides a complementary basis for characterizing concentrated gas hydrate deposits, as the long-offset data allow retrieving the P-wave velocity information of the subsurface, whereas the higher resolution data enable detailed imaging of geologic features associated with gas hydrates and fluid flow.
Chapter 3 focuses on two elongated four-way closures systems at the toe of the Hikurangi deformation wedge - Glendhu and Honeycomb ridges, which host hydrate deposits in high concentration. I found that the mechanism for concentrated hydrate accumulation is along-strata gas migration, and that the vertical extent of these accumulations is a function of the steepness of the strata crossing the base of the gas hydrate stability zone and of the volume of sediments from which fluid flows into each structure.
In Chapter 4, I present a quantitative assessment of the total gas hydrate in place in these deposits, by carrying out a two-step inversion of long-offset seismic profiles that cross areas of concentrated hydrate deposits at Glendhu and Honeycomb ridges (APB13-21 and APB13-17). First, the elastic properties of the subsurface were estimated through geostatistical amplitude-versus-angle seismic inversion. Secondly, these properties were inverted for porosity and gas hydrate saturation through a Bayesian framed petrophysical inversion built on a rock physics model for hydrate-bearing marine sediments. This geological modelling workflow allowed retrieval of the spatial distribution of gas hydrate saturation and porosity in the gas hydrate stability zone. Among the advantages of combining geostatistical AVA seismic inversion with a Bayesian rock physics inversion is the possibility to access the variability associated with the results. Estimates of the total gas hydrate volume in place at these two reservoirs are between 2.45 x 10^5 m^3 and 1.72 x 10^6 m^3, with the best estimate at 9.68 x 10^5 m^3.
A more detailed characterization of the gas hydrate stability zone was achieved through 1D full-waveform inversion. The goal of this study, presented in Chapter 5, was to retrieve the fine-scale P-wave velocity profiles at selected locations corresponding to specific structures related to gas hydrates and fluid flow observed in the seismic data. The results show high velocity anomalies associated with high reflectivity with positive polarity, interpreted as porous units bearing gas hydrate in high concentration. Lower velocities are predicted in correspondence of chimney-like structures, which could possibly be associated with focused fluid flow. However, the resolution of the inverted velocity models depends on the vertical resolution of the seismic data, which could be at least an order of magnitude larger than the thickness of interbedded sandy and silty layers of typical lower wedge sedimentary units.
Finally, to investigate the large-scale methane seepage from five known sites of gas venting, I integrate the seismic data with hydroacoustic data from both multibeam and split-beam echo sounders collected over three years. Qualitative analysis of multibeam data, which allowed mapping the backscatter anomalies near the seafloor at the sites of seepage, suggests that there is no substantial difference in the activity of the main seeps between 2018 and 2020. Moreover, the use of the multi-frequency split-beam echosounder allowed estimates of the gas flux rates at the five target sites to be made. The cold seeps analysed in Chapter \ref{chap:chapter6} lie in water depths ranging from 1110 to 2060 m, and emit, combined, between 9.52 and 27.52 x 10^6 kg of gas per year. This study provides a quantitative assessment of the greenhouse gases contribution from the southern Hikurangi Margin, and has implications at local and global scale.