1989 Theses Doctoral

Statistical studies of seafloor geomorphology

Malinverno, Alberto

In general, topography can be subdivided in two components.

The first component can be named deterministic, and it is that part of the topographic signal that can be explained in terms of simple, predictive physical models. An example is the progressive deepening of the ocean floor as it moves away from a mid-ocean ridge, which is due to the progressive cooling, contraction and isostatic subsidence of the oceanic lithosphere. The second component is here named stochastic, and it is that part of the topographic signal that varies randomly and is essentially unpredictable in detail. An example is the rough abyssal hilltopography, created by time-varying volcanism and faulting and superimposed on the smooth subsidence of mid-ocean ridge flanks.

While most geophysical studies of topography concentrate on some deterministic component, this thesis is a study of the stochastic component of seafloor topography. As this component varies randomly, a number of statistical approaches are used to parameterize the variation of the stochastic component of topography.

In the first chapter of this thesis, profiles of topography are examined as traces of random noise, and the data are compared with the predictions of two simple statistical models (a self-similar and an autoregressive model). Chapter 2 is an attempt to deal with non-stationarity, i.e., with the fact that the statistical properties of topography vary in space. Chapter 3 presents a simple model for the generation of ridge flank, abyssal hill topography by faulting: a series of random faults is convolved with an impulse response to faulting to give as output the observed topography. Chapter 4 starts from the assumption that each hill corresponds to a volcanic or tectonic event, and shows how sequences of hills can be treated as an aperiodic Poisson process.

Finally, chapter 5 is a study of the along-strike variations in the axial topography of the Mid-Atlantic Ridge and its relationship with corresponding variations in spreading rate, distance from and offset on neighboring transforms, and depth of the zero-age lithosphere.