Search

Time-lapse images carry out important information about dynamic changes in Earth's interior which can be inferred using different Full Waveform Inversion (FWI) schemes. The estimation process is performed by manipulating more than one seismic dataset, associated with the baseline and monitors surveys. The time-lapse variations can be so minute and localised that quantifying the uncertainties becomes fundamental to assessing the reliability of the results. The Bayesian formulation of the FWI problem naturally provides confidence levels in the solution, but evaluating the uncertainty of time-lapse seismic inversion remains a challenge due to the ill-posedness and high dimensionality of the problem. The Hamiltonian Monte Carlo (HMC) can be used to effectively sample over high dimensional distributions with affordable computational efforts. In this context, we propose a probabilistic Bayesian sequential approach for time-lapse FWI using the HMC method. Our approach relies on the integration of the baseline survey information as prior knowledge in the monitor estimation. We compare the proposed methodology with a parallel scheme in perfect and perturbed acquisition geometry scenarios. We also investigate the correlation effect between baseline and monitor samples in the propagated uncertainties. The results show that our strategy provides accurate times-lapse estimates with errors of similar magnitude to the parallel methodology., Comment: 12 pages, 6 figures

Full-Waveform Inversion (FWI) is a high-resolution technique used in geophysics to evaluate the physical parameters and construct subsurface models in a noisy and limited data scenario. The ill-posed nature of the FWI turns this a challenging problem since more than one model can match the observations. In a probabilistic way, solving the FWI problem demands efficient sampling techniques to infer information on parameters and to estimate the uncertainties in high-dimensional model spaces. We investigate the feasibility of applying the Hamiltonian Monte Carlo (HMC) method in the acoustic FWI by a reflection setup containing different noise level data. We propose a new strategy for tuning the mass matrix based on the acquisition geometry of the seismic survey. Our methodology significantly improves the ability of the HMC method in reconstructing reasonable seismic models with affordable computational efforts., Comment: 10 pages, 6 figures

The conventional approach to data-driven inversion framework is based on Gaussian statistics that presents serious difficulties, especially in the presence of outliers in the measurements. In this work, we present maximum likelihood estimators associated with generalized Gaussian distributions in the context of R\'enyi, Tsallis and Kaniadakis statistics. In this regard, we analytically analyse the outlier-resistance of each proposal through the so-called influence function. In this way, we formulate inverse problems by constructing objective functions linked to the maximum likelihood estimators. To demonstrate the robustness of the generalized methodologies, we consider an important geophysical inverse problem with high noisy data with spikes. The results reveal that the best data inversion performance occurs when the entropic index from each generalized statistic is associated with objective functions proportional to the inverse of the error amplitude. We argue that in such a limit the three approaches are resistant to outliers and are also equivalent, which suggests a lower computational cost for the inversion process due to the reduction of numerical simulations to be performed and the fast convergence of the optimization process., Comment: 17 pages, 14 figures

In this work we propose a robust methodology to mitigate the undesirable effects caused by outliers to generate reliable physical models. In this way, we formulate the inverse problems theory in the context of Kaniadakis statistical mechanics (or $\kappa$-statistics), in which the classical approach is a particular case. In this regard, the errors are assumed to be distributed according to a finite-variance $\kappa$-generalized Gaussian distribution. Based on the probabilistic maximum-likelihood method we derive a $\kappa$-objective function associated with the finite-variance $\kappa$-Gaussian distribution. To demonstrate our proposal's outlier-resistance, we analyze the robustness properties of the $\kappa$-objective function with help of the so-called influence function. In this regard, we discuss the role of the entropic index ($\kappa$) associated with the Kaniadakis $\kappa$-entropy in the effectiveness in inferring physical parameters by using strongly noisy data. In this way, we consider a classical geophysical data-inverse problem in two realistic circumstances, in which the first one refers to study the sensibility of our proposal to uncertainties in the input parameters, and the second is devoted to the inversion of a seismic data set contaminated by outliers. The results reveal an optimum $\kappa$-value at the limit $\kappa \rightarrow 2/3$, which is related to the best results., Comment: 37 pages, 23 figures

We show that the same maximum entropy principle applied to recurrence microstates configures a new way to properly compute the time delay necessary to correctly sample a data set. The new method retrieves results obtained using traditional methods with the advantage of being independent of any free parameter. Since all parameters are automatically set, the method is suitable for use in artificial (computational) intelligence algorithms, recovering correct information embedded in time series, and rationalizing the process of data acquisition since only relevant data must be collected., Comment: 5 pages, 4 figures

Recurrence entropy $(\cal S)$ is a novel time series complexity quantifier based on recurrence microstates. Here we show that $\mathsf{max}(\cal S)$ is a \textit{parameter-free} quantifier of time correlation of stochastic and chaotic signals, at the same time that it evaluates property changes of the probability distribution function (PDF) of the entire data set. $\mathsf{max}(\cal S)$ can distinguish distinct temporal correlations of stochastic signals following a power-law spectrum, $\displaystyle P(f) \propto 1/f^\alpha$ even when shuffled versions of the signals are used. Such behavior is related to its ability to quantify distinct subsets embedded in a time series. Applied to a deterministic system, the method brings new evidence about attractor properties and the degree of chaoticity. The development of a new parameter-free quantifier of stochastic and chaotic time series opens new perspectives to stochastic data and deterministic time series analyses and may find applications in many areas of science.

Barkhausen effect in ferromagnetic materials provides an excellent area for investigating scaling phenomena found in disordered systems exhibiting crackling noise. The critical dynamics is characterized by random pulses or avalanches with scale-invariant properties, power-law distributions, and universal features. However, the traditional Barkhausen avalanches statistics may not be sufficient to fully characterize the complex temporal correlation of the magnetic domain walls dynamics. Here we go beyond power laws and focus on the multifractal scenario to quantify the temporal scaling characteristics of Barkhausen avalanches in polycrystalline and amorphous ferromagnetic films with thicknesses from $50$ nm to $1000$ nm. We show that the multifractal properties are dependent on the film thickness, and insensitive to the structural character of the materials. Further, we observe for the first time the collapse of the multifractality in the domain walls dynamics as the thickness is reduced, and the multifractal behavior gives place to a monofractal one over the entire range of time scales. The reorganization in the temporal scaling characteristics of Barkhausen avalanches is understood as an universal restructuring associated to the dimensional transition, from three to two-dimensional magnetization dynamics., Comment: 3 figures

In this letter we explore how recurrence quantifier, the determinism ($\Delta$), can reveal stationarity breaking and coupling between physical systems. We demonstrate that it is possible to detect small variations in a dynamical system based only on temporal signal displayed by another system coupled to it. To introduce basic ideas, we consider a well known dynamical system composed of two master-slave coupled Lorenz oscillators. We start evidencing that due to the sensitivity of $\Delta$ computed from temporal time series of slave oscillator, its is possible to detect the stationary breaking imposed in the master oscillator. As a second example, the method is carried out in a real physiological data acquired from accelerometer sensors ($\mathrm{A_{cc}}$) and used to detect micro arousal phenomenology (described by a sharp burst in $\mathrm{A_{cc}}$ signal) during sleep periods in mice. Moreover, we show for the first time that making use of recurrence quantifier it is possible to infer a coupling between electric signals from hippocampus and "locomotor brain areas" of mice, based only on non invasive data from $\mathrm{A_{cc}}$. Our results suggest new possibilities of analysis of coupled systems making use of accessible time series. Our second example supports an interpretation of an internal coupling detectable as a stationarity breaking in $\mathrm{A_{cc}}$ that occurs some seconds before micro arousal processes during sleep periods in rodents, contributing to the idea that micro arousals are elements of sleep taking part in the regulation of sleep process. Such a characterization of micro arousals can improve our knowledge about sleep fostering tools of sleep diagnose and pharmacology research for mammals in general., Comment: After the review process, the authors do not agree that the paper should be submitted to the arXv. So no future version of the text will be uploaded

A recent problem in community ecology lies in defining structures behind matrices of species interactions. The interest in this area is to quantify the nestedness degree of the matrix after its maximal packing. In this work we evaluate nestedness using the sum of all distances of the occupied sites to the vertex of the matrix. We calculate the distance for two artificial matrices with the same size and occupancy: a random matrix and a perfect nested one. Using these two benchmarks we develop a nestedness estimator. The estimator is applied to a set of 23 real networks of insect-plant interactions., Comment: 16 pages, 8 figures, 1 table

We develop a network in which the natural numbers are the vertices. We use the decomposition of natural numbers by prime numbers to establish the connections. We perform data collapse and show that the degree distribution of these networks scale linearly with the number of vertices. We compare the average distance of the network and the clustering coefficient with the distance and clustering coefficient of the corresponding random graph. In case we set connections among vertices each time the numbers share a common prime number the network is not a small-world type. If the criterium for establishing links becomes more selective, only prime numbers greater than $p_l$ are used to establish links, the network shows small-world effect, it means, it has high clustering coefficient and low distance.