Carlos Francisco Castro Soto Reyes

A thesis for the degree of Doctor of Philosophy  defended February 2019.

The PhD School of Science, Faculty of Science, Physics of Ice Climate and Earth, Niels Bohr Institute, University of Copenhagen

Supervisor:
Prof. Klaus Mosegaard

The crustal and sedimentary structure of the Amundsen Basin, Arctic Ocean, derived from seismic reflection and refraction data

The Amundsen Basin is the deepest abyssal plain in the Arctic Ocean that separates the  continental Lomonosov Ridge from the Gakkel Ridge, the current seafloor spreading axis in the  Eurasian Basin. The basin was created by ultraslow seafloor spreading at the Gakkel Ridge and consists of alternating magmatic and amagmatic ridge segments; however, it is unclear if this is true for the entire opening history of the basin. The sedimentary history of the basin is still poorly constrained due to perennial sea ice cover and the associated logistical challenges of acquiring geophysical, geological and in particular also well data. This dissertation analyses one of the few geophysical data sets available for the western Amundsen Basin to improve the understanding of both the stratigraphic and the crustal accretional history of the basin.From 2007 to 2012, three expeditions (LOMROG I through III) were carried out to acquire seismic data in the western Amundsen Basin in the Arctic Ocean.The data of the LOMROG expeditions consist of 1028 km of seismic reflection data and 118 sonobuoys deployed along the seismic lines to obtain information on the velocity structure of the sediments and crust. For the analysis and interpretation, additional information was used including published multichannel seismic data in the Amundsen Basin, refraction seismic data, magnetic data, gravity data, and the wells of the Arctic Coring Expedition (ACEX).

The seismic refraction data were used to develop P-wave velocity models for the sediments, crust, and uppermost mantle utilizing forward modeling techniques of travel times. The initial geometry of the sediment layers in the models were made in combination with the coincident multichannel seismic data, allowing to add more detail down to the basement than what would have been possible with refraction data alone. The multichannel seismic data were used to develop a stratigraphic model based on the reflection character, seismic facies, and geometries of each stratigraphic unit. Once both the velocity and the stratigraphic models were complete, sedimentation rates were calculated for each unit based on the velocities obtained from the refraction data, two-way travel times in the seismic reflection data, and age constraints from magnetic data and known tectonic and oceanographic events in the Arctic Ocean. The sedimentation rates were then used to infer possible depositional environments within the basin’s history.

The seismic stratigraphy analysis places new constraints on the Cenozoic depositional history. Four distinct phases of basin development are recognized. From the onset of spreading up to the mid-Oligocene, a small, isolated basin dominated by processes that are tectonically controlled is indicated. During the late Oligocene to early Miocene, widespread passive infill associated with hemipelagic deposition reflects a phase of tectonic quiescence, most likely in a freshwater estuarine setting. During the middle Miocene, mounded sedimentary build-ups along the Lomonosov Ridge suggest the onset of geostrophic bottom-currents that likely formed in response to a deepening and widening of the Fram Strait. In contrast, the Plio–Pleistocene stage is characterized by erosional features such as scarps and channels adjacent to levee accumulations, indicative of a change to a higher-energy environment. These deposits are suggested to be partly associated with dense shelf water-mass plumes driven by supercooling and brine formation originating below thick multi-year sea-ice over the northern Greenland continental shelf.

P-wave modeling of the crust and upper mantle was supplemented with gravity modeling in order to determine the Moho depth in areas with low seismic resolution. The velocity models reveal a detailed picture of the crustal velocity structure of the basin. Three distinct basement types are identified: oceanic crust with layers 2 and 3, oceanic crust with a layer 3 that is absent, and an exhumed and serpentinized mantle. The total maximum observed thickness in the basin is 6 km but typically ranges between 2–5 km. Moreover, the seismic modeling indicates the presence of velocities compatible with an oceanic layer 2 and 3 within the extensions of the amagmatic sector of the Gakkel Ridge. These results are different than previous observations along the Gakkel Ridge, where no oceanic layer 3 has been documented. The different basement types therefore indicate that there exists both a spatial and temporal variation in crustal accretion processes at the ridge.