Produção Científica
Artigo em Revista
Parallel Source Scanning Algorithm using GPUs The use of methods using waveform stacking are nowadays more common in microseismic monitoring applications because they avoid manual or automatic phase picking. The Source Scanning Algorithm (SSA) is a widely known technique in which the source location is estimated using a brightness function obtained from stacking the normalized absolute amplitude seismograms recorded at several stations. The SSA has the advantage of the straightforwardness of its implementation but has the inconvenience of being computationally costly even for smallscale experiments. Our approach is then to parallelize the sequential SSA using graphics processing units (GPUs), and we named this parallel version pSSA. We have parallelized the Stacking step of the SSA Algorithm because this is by far the most computationally demanding Step. This can be done efficiently because of the spatial independence of the data. In our test cases we performed sequential and parallel computations of the SSA and pSSA in two different platforms. Additionally, we compared the performance of pSSA with a parallel implementation using OpenMP. We demonstrate that pSSA has produced speedups up to 125Ã— as compared to the sequential version. We implemented a clientâ€“server architecture to receive and process the data. This architecture can treat with various simultaneous clients and also with outoforder data packets. This allows for resending lost or corrupted data. We anticipate that pSSA has the impact of allowing SSA like algorithm to be used in microseismic experiment design and the use of onsite realtime denoising techniques, as well as the potential of being used in traffic light warning systems for fluid injection operations. 

Artigo em Revista
Karstification and fluid flow in carbonate units controlled by propagation and linkage of mesoscale fractures, JandaÃra Formation, Brazil This study examines how centimeter to meterlong fracture sets propagate, link and form mlong structures, which are then karstified in the vadose (aerated) zone in the upper parts of exposed layered carbonate units. We characterize the fracture patterns, including both mesoscale extensional joints, veins, stylolites and macroscale faults, within the Rosario pavement, which is a 1100mlong, 340mwide outcrop in the JandaÃra Formation, Potiguar Basin in semiarid Brazil. The work compares the directions of the mesoscale structures with the strikes of surface collapse dolines and cave conduits at depths of 10â€“20 m. The main results indicate that the background (diffuse) deformation consists mostly of bedperpendicular, stratabound NSstriking veins and extensional (opening modeI) fractures and EW to ENEWSWstriking stylolites, which are consistent with the same stress field. Both sets of structures are laterally continuous and are commonly traceable across the pavement. Linkage of stylolites and stratabound fractures formed via throughgoing fractures that affected several layers, as deep as 15 m. The throughgoing fractures were favorable structures for fluid pathway formation and karstification in the epigenetic environment. We conclude that sets of mesoscale structures fit with the orientations of the collapse dolines and subhorizontal cave passages, and induced hydraulic anisotropy within the vadose level. These findings indicate that the fractures, veins and stylolites grow, link and form localized vertical conduits for water percolation and karst development. These findings can constrain future models and numerical simulations of karst conduits in investigations of groundwater and oil reservoirs. 

Artigo em Revista
Can Euler deconvolution outline threedimensional magnetic sources? Severe limitations of the standard Euler deconvolution to outline source shapes have been pointed out. However, Euler deconvolution has been widely employed on field data to outline interfaces, as faults and thrust zones. We investigate the limitations of the 3D Euler deconvolutionâ€“derived estimates of source dip and volume with the use of reducedtothepole synthetic and field anomalies. The synthetic anomalies are generated by two types of source bodies: (1) uniformly magnetized prisms, presenting either smooth or rough interfaces, and (2) bodies presenting smooth delimiting interfaces but strong internal variation of magnetization intensity. The dip of the first type of body might be estimated from the Euler econvolution solution cluster if the ratio between the depth to the top and vertical extent is relatively high (>1/4). For the second type of body, besides dip, the source volume can be approximately delimited from the solution cluster envelope, regardless of the referred ratio. We apply Euler deconvolution to two field anomalies which are caused by a curvedshape thrust zone and by a banded iron formation. These anomalies are chosen because they share characteristics with the two types of synthetic bodies. For the thrust zone, the obtained Euler deconvolution solutions show spatial distribution allowing to estimate a source dip that is consistent with the surface geology data, even if the abovementioned ratio is much less than 1/4. Thus, there are other factors, such as a heterogeneous magnetization, which might be controlling the vertical spreading of the Euler deconvolution solutions in the thrust zone. On the other hand, for the ironore formation, the solution cluster spreads out occupying a volume, in accordance with the results obtained with the synthetic sources having internal variation of magnetization intensity. As conclusion, although Euler deconvolutionâ€“derived solutions cannot offer accurate estimates of source shapes, they might provide a sufficient degree of reliability in the initial estimates of the source dip and volume, which may be useful in a later phase of more accurate modelling. 

Artigo em Revista
A new look at the causes of â€œpolarizationâ€ horns in electromagnetic well logging Socalled polarization horns appear in electromagnetic well logs when the sondes move through the interface between formations of different conductivities. Since the early 1990s, peaks in the logs have been attributed to the influence of surface charges due to the presence of electric field components perpendicular to the interfaces. A new analysis finds that surface charges should not be pinpointed as the immediate cause for the appearance of interface peaks in electromagnetic logs from any tool operating in any frequency range. This is accomplished by calculating the derivatives of the magnetic field components with respect to the sondeâ€™s vertical position as it crosses the interface between two homogeneous isotropic media. The mathematical expressions reveal the components with smooth transitions and the ones with discontinuities in their rates of change across the interface. The analysis is applied to the four components needed to simulate the responses of coaxial and coplanar coil configurations in dipping logs. The results show that only the horizontal field from a horizontal dipole source suffers an influence from the current density field perpendicular to the interface between the two media, which gives rise to surface charges, but even for this component, the nonsmooth transitions are not associated with the perpendicular current. An anisotropic example gives further support to the conclusion that the polarization horns are associated with the discontinuous current density field parallel to the interface rather than with the continuous current across the interface. 

Artigo em Revista
GEMM3D: An Edge Finite Element program for 3D modeling of electromagnetic fields and sensitivities for geophysical applications This paper presents the 3D modeling program GEMM3D (Geophysical ElectroMagnetic Modeling 3D), a Fortran program for modeling frequency domain electromagnetic responses of geophysical sources in threedimensional models. The program is an implementation of the Edge Finite Element method, in an electric field formulation, to simulate any electromagnetic source of interest, at all frequencies that are used in the geophysical methods in the quasistatic regime. Its current version includes routines for plane waves and simulation of magnetotelluric data; electrical and magnetic dipoles; circular loops. Other sources and configurations may easily be included with the addition of routines for simulating primary (layered earth) fields. The program includes routines for calculating sensitivities using the Adjoint State method, which makes efficient use of direct solvers for factoring once and solving for multiple systems with the same coefficient matrix. The paper details the methods, the implementation, and the usage of the program. Validation tests are presented for different sources. The program is freely available for use in research and teaching. 

Artigo em Revista
Analysis of EshelbyChengâ€™s model in anisotropic porous cracked medium: An ultrasonic physical modeling approach Many effective medium theories are designed to describe the macroscopic properties of a medium (the rock, or reservoir in this case) in terms of the properties of its constituents (the background matrix of the rock and the inclusions, for our scenario). A very well known effective medium theory is the EshelbyCheng model, which was studied by us in previous work, being tested for the case where the background medium was weaklyanisotropic and porous. The analysis was done testing elastic velocities and Thomsen parameters  as a function of crack density for fixed values of aspect ratio  predicted by the model with data acquired from synthetic rock samples. In this work, we aim to complete the analysis of the EshelbyCheng model capabilities when applied to rocks with porous and vertical transversely isotropic (VTI) backgrounds, testing the model for the elastic velocities as functions of aspect ratio  for fixed values of crack density  against experimental data. The data used to test the model were obtained from 17 synthetic rock samples, one uncracked and 16 cracked, the latter divided into four groups of four samples each, each group with cracks having the same aspect ratio, but with the samples having different crack densities. In these samples, ultrasonic pulse transmission measurements were performed to obtain the experimental velocities used to test the model. As was not possible to acquire data for velocity as a function of aspect ratio for fixed values of crack density, we performed interpolations of the experimental data to estimate these velocities. EshelbyCheng model effective velocities and Thomsen parameters were calculated using three formulations proposed for the crack porosity: one proposed by Thomsen, the second one proposed in our previous work (which depends only on the crack density) and the third one proposed in this work (which depends on the crack porosity and the aspect ratio, just like Thomsenâ€™s proposal). The comparisons between elastic velocities and Thomsen parameters  as function of crack aspect ratio, for fixed values of crack density  predicted by the model and estimated from the data via interpolation showed that the third formulation produced better fittings (lower rootmeansquare errors) between model and experimental data for all ranges of aspect ratio and crack density. 

Artigo em Revista
On the seismic wavelet estimative and reflectivity recovering based on linear inversion: Welltoseismic tie on a real data set from Viking Graben, North Sea Tying seismic data to well data is critical in reservoir characterization. In general, the main factors controlling a successful seismic well tie are an accurate timedepth relationship and a coherent wavelet estimate. Wavelet estimation methods are divided into two major groups: statistical and deterministic. Deterministic methods are based on using the seismic trace and the well data to estimate the wavelet. Statistical methods use only the seismic trace and generally require assumptions about the waveletâ€™s phase or a random process reflectivity series. We have compared the estimation of the wavelet for seismic well tie purposes through leastsquares minimization and zeroorder quadratic regularization with the results obtained from homomorphic deconvolution. Both methods make no assumption regarding the waveletâ€™s phase or the reflectivity. The bestestimated wavelet is used as the input to sparsespike deconvolution to recover the reflectivity near the well location. The results show that the wavelets estimated from both deconvolutions are similar, which builds our confidence in their accuracy. The reflectivity of the seismic section is recovered according to known stratigraphic markers (from gammaray logs) resent in the real data set from the Viking Graben field, Norway. 

Artigo em Revista
Interval mineral and fluid densities estimation from welllogs: Application to the Norne Field dataset Formation evaluation techniques are the key to understand subsurface rocks properties from welllogs, especially those drilled in hydrocarbon exploration wells. Knowledge of the parameters related to different types of rocks is traditionally used in forward determination of lithology, porosity and water saturation, which can be refined by calibrating the input models. In this work, we perform welllog interval linear inversion with respect to formation density to investigate mineral and fluid properties in a real dataset. The method is based on an overdetermined problem, which supposes a homogeneous distribution of petrophysical parameters through stratigraphic layers and is applied in conventional reservoir rocks from the Norne Field (offshore Norway). Bulk density, gammaray and neutron porosity logs are employed in a workflow that relies on layerbylayer leastsquares regressions to estimate matrix, shale and fluid apparent densities. In this process, shale volume and the total density porosity are calculated depth by depth from empirical equations feed by the input logs. Therewith, the introduced inversion scheme stands as an alternative approach for well data interpretation that focuses on computing the densities of the rock constituents instead of fixing these parameters to invert fractional volumes. Furthermore, the application in two wellbores resulted in geological consistent individual densities in most intervals, except for a gasbearing zone observed in one of the boreholes, where porosity uncertainty caused anomalous variationin grain and fluid densities. 

Artigo em Revista
Seismic Stratigraphy of Camamu Basin, Northeastern Brazil â€”The Camamu Basin is located at the northeastern Brazilian coastline and has significant hydrocarbon potential in both shallow and deep water settings. However, despite an already operating productive gas field, the basin is not well known. Herein, a regional stratigraphic interpretation of the rift, the transitional and the drift megasequences is reported, based on a data set of 152 poststacked 2D seismic lines, 1 3D seismic cube and 34 wells with eletrologs. The study revealed that the rift megasequence is much more complex than previously thought. In the southern region of the Camamu Basin, five rift sequences were mapped, while in the northern region, in the area of the regional tectonic lineament known as the Salvador Transcurrent Zone (STZ), seven rift sequences were recognized. This difference suggests a tectonic control during the rifting process, because the northern region is intensely affected by shear stress induced by transcurrent tectonics of the STZ during the crustal breakup. For the postrift or â€˜â€˜transitionalâ€™â€™ phase associated with thermal subsidence installed after the rift, as well as for the drift succession, the tectonic control exerted by Salvadorâ€™s Transcurrent Zone is not detectable and the tectonosedimentary evolution of the basin follows the general pattern of the Brazilian marginal basins. 
Artigo em Revista
NonHyperbolic Velocity Analysis of Seismic Data from Jequitinhonha Basin, Northeastern Brazil Normal moveout (NMO) velocity is used in seismic data processing to correct the data from the moveout effect. This velocity depends on the medium above the reflector and it is estimated from the adjustment of a hyperbolic function that approximates the reflection time. This approximation is reasonable for media formed by isotropic layers. For deeper exploration targets, which effectively behave as anisotropic media,the NMO velocity estimate from the hyperbolic approximation becomes imprecise. One possibility is the use of nonhyperbolic approximations for the reflection time and deeming the medium to be anisotropic. However, these approximations make the NMO velocity estimation a more complex problem, since the anisotropic parameters are unknown. In this study the NMO velocities for a vertical transverse isotropy medium are estimated using two nonhyperbolic reflection time approaches. For comparing the two methodologies that estimate NMO velocity, a 2D dataset from Jequitinhonha Basin is used and it presents anisotropic behavior. The results show that this approach produces more consistent results than the conventional approach, which ignores the anisotropy of the medium. 
