Turning noise into useful data-every geophysicist's dream? And now it seems possible. The field of seismic interferometry has at its foundation a shift in the way we think about the parts of the signal that are currently filtered out of most analyses-complicated seismic codas (the multiply scattered parts of seismic waveforms) and background noise (whatever is recorded when no identifiable active source is emitting, and which is superimposed on all recorded data). Those parts of seismograms consist of waves that reflect and refract around exactly the same subsurface heterogeneities as waves excited by active sources. The key to the rapid emergence of this field of research is our new understanding of how to unravel that subsurface information from these relatively complex-looking waveforms. And the answer turned out to be rather simple. This article explains the operation of seismic interferometry and provides a few examples of its application.A simple thought experiment. Consider an example of a horizontally stratified (one-dimensional) acoustic medium, and for the moment let us imagine that it has only a single internal interface. Now, say horizontally planar pressure waves are emitted by two impulsive sources, one after the other, and that one source is above the interface and one below. Vibrations from the resulting propagating waves are recorded at two receivers which can be placed anywhere between the two sources ( Figure 1, left).The recordings are shown in the center of the figure. At each receiver a direct and a reflected wave is recorded for source 1, whereas only one transmitted wave is recorded for source 2.Seismic interferometry of these data involves only two simple steps: The two recorded signals from each source are crosscorrelated and the resulting crosscorrelograms are summed (stacked). The result, shown on the right of Figure 1, is surprising; for positive times it is the seismogram that would have been recorded at either receiver if the other receiver had in fact been a source, and at negative times it is the time reverse of this seismogram. In other words, by this simple, two-step operation we have constructed the seismic trace from a virtual source-a source that did not exist in our initial experiment, and a source that is imagined to be at the location of one of our receivers.To generalize, this simple example placed no constraint on where the receivers were placed, provided they were between the sources. By moving either or both of them (or by using many distributed receivers from the start), it is therefore possible to construct the trace from an infinite number of virtual source and receiver pairs placed at any locations, by recording the signal from only two actual sources. What is more, provided one of the active sources is above the interface and receivers and the other is below, the location of the active sources is also arbitrary, and in order to carry out the process above we do not even need to know where these sources are. Seismic interferometry steps.The fundamental steps of t...

A c c e p t e d M a n u s c r i p t 3 ABSTRACTThe first appearance of skeletal metazoans in the late Ediacaran (~550 million years ago; Ma) has been linked to the widespread development of oxygenated oceanic conditions, but a precise spatial and temporal reconstruction of their evolution has not been resolved. Here we consider the evolution of ocean chemistry from ~550 to ~541 Ma across shelf-to-basin transects in the Zaris and Witputs Sub-Basins of the Nama Group, Namibia. New carbon isotope data capture the final stages of the Shuram/Wonoka deep negative C-isotope excursion, and these are complemented with a reconstruction of water column redox dynamics utilizing Fe-S-C systematics and the distribution of skeletal and soft-bodied metazoans. Combined, these inter-basinal datasets provide insight into the potential role of ocean redox chemistry during this pivotal interval of major biological innovation.The strongly negative  13 C values in the lower parts of the sections reflect both a secular, global change in the C-isotopic composition of Ediacaran seawater, as well as the influence of 'local' basinal effects as shown by the most negative  13 C values occurring in the transition from distal to proximal ramp settings. Critical, though, is that the transition to positive  13 C values postdates the appearance of calcified metazoans, indicating that the onset of biomineralization did not occur under post-excursion conditions. Significantly, we find that anoxic and ferruginous deeper water column conditions were prevalent during and after the transition to positive  13 C that marks the end of the Shuram/Wonoka excursion. Thus, if the C isotope trend reflects the transition to global-scale oxygenation in the aftermath of the oxidation of a large-scale, isotopically light organic carbon pool, it was not sufficient to fully oxygenate the deep ocean. Page 4 of 74A c c e p t e d M a n u s c r i p t 4 Both sub-basins reveal highly dynamic redox structures, where shallow, inner ramp settings experienced transient oxygenation. Anoxic conditions were caused either by episodic upwelling of deeper anoxic waters or higher rates of productivity. These settings supported short-lived and monospecific skeletal metazoan communities. By contrast, microbial (thrombolite) reefs, found in deeper inner-and mid-ramp settings, supported more biodiverse communities with complex ecologies and large skeletal metazoans. These long-lived reef communities, as well as Ediacaran soft-bodied biotas, are found particularly within transgressive systems, where oxygenation was persistent. We suggest that a mid-ramp position enabled physical ventilation mechanisms for shallow water column oxygenation to operate during flooding and transgressive sea-level rise. Our data support a prominent role for oxygen, and for stable oxygenated conditions in particular, in controlling both the distribution and ecology of Ediacaran skeletal metazoan communities.Keywords: Oxygenation; Neoproterozoic; Biomineralisation; Metazoans; Ediacaran; Ecosystems Introductio...

S U M M A R YWe present a neural network approach to invert surface wave data for a global model of crustal thickness with corresponding uncertainties. We model the a posteriori probability distribution of Moho depth as a mixture of Gaussians and let the various parameters of the mixture model be given by the outputs of a conventional neural network. We show how such a network can be trained on a set of random samples to give a continuous approximation to the inverse relation in a compact and computationally efficient form. The trained networks are applied to real data consisting of fundamental mode Love and Rayleigh phase and group velocity maps. For each inversion, performed on a 2 • × 2 • grid globally, we obtain the a posteriori probability distribution of Moho depth. From this distribution any desired statistic such as mean and variance can be computed. The obtained results are compared with current knowledge of crustal structure. Generally our results are in good agreement with other crustal models. However in certain regions such as central Africa and the backarc of the Rocky Mountains we observe a thinner crust than the other models propose. We also see evidence for thickening of oceanic crust with increasing age. In applications, characterized by repeated inversion of similar data, the neural network approach proves to be very efficient. In particular, the speed of the individual inversions and the possibility of modelling the whole a posteriori probability distribution of the model parameters make neural networks a promising tool in seismic tomography.

SUMMARYWe present the first Love-wave group velocity and shear velocity maps of the British Isles obtained from ambient noise interferometry and fully non-linear inversion. We computed interferometric inter-station Green's functions by cross-correlating the transverse component of ambient noise records retrieved by 61 seismic stations across the UK and Ireland. Group velocity measurements along each possible inter-station path were obtained using frequency-time analysis and converted into a series of inter-station traveltime datasets between 4 and 15 seconds period. Traveltime uncertainties estimated from the standard deviation of dispersion curves constructed by stacking randomly-selected subsets of daily cross-correlations, were observed to be too low to allow reasonable data fits to be obtained during tomography. Data uncertainties were therefore estimated again during the inversion as distance-dependent functionals. We produced Love-wave group velocity maps within 8 different period bands using a fully non-linear tomography method which combines the transdimensional reversible-jump Markov chain Monte Carlo (rj-McMC) algorithm with an

S U M M A R YSeismic interferometry can be used to estimate interreceiver surface wave signals by crosscorrelation of signals recorded at each receiver. The quality of the estimated surface waves is controlled by the distribution of sources exciting the cross-correlated wavefields, and it is commonly thought that only sources at or near the surface are required to generate accurate estimates. We study the role of source distribution in surface wave interferometry for both surface and subsurface sources using surface wave Green's functions for laterally homogeneous media. We solve the interferometric integral using a Rayleigh wave orthogonality relationship combined with a stationary phase approach. Contrary to popular opinion we find that sources at depth do indeed play a role in the recovery of surface waves by interferometry. We find that interferometry performs well when surface sources are distributed homogeneously at the surface of the Earth. However, when this homogeneous distribution is not available amplitude errors are introduced, and when multiple modes are present strong spurious events appear and higher mode surface waves may not be correctly estimated. In order to recover higher mode surface waves we propose an additional step in the processing of surface wave data for seismic interferometry: by separating modes and applying interferometry to each mode individually it is possible to recover the interreceiver surface wave modes, without the artefacts introduced by limited source coverage.