seisFacies When there are few wells available for reservoir description, the seismic attributes can be used to study stratigraphic facies. The sliced attribute/property, which can be generated from our GeoInversion system, is new geologic time surface by proportionally slicing between reference seismic horizons. And the sliced attribute/property will significantly deepen your understanding of your stratigraphic facies and your depositional history of your reservoir. A self-organization neural network method will be used to cluster the attributes into several groups which represent different stratigraphic facies. The existed well facies will guide the interpretations and evaluations.

Vp/Vs/Den Inversion (seisInversion): There are two methods for seismic inversion. The first method is (non-)Linear Conjugate Gradient method to estimate the Vp, Vs and density. Another method is Bayesian-based Support Vector Machine method to incorporate the petrophysical property and seismic attributes, which has high resolution with similar to well logs. The Vp/Vs/Den can be used to calculate the Young's modulus, Poisson Ratio, UCS, Pore Pressure and minimum horizontal stress, which will be useful to wellbore stability analysis and shale gas reservoir evaluation.

Seismic Geological Modeling (seisSuperMan) An advanced horizon and fault manipulation tools, including interactive picking of target horizons and faults. The tool also supports to load seismic horizons from other system. The target layers, which build the geological models, compose of top/bottom horizons and left/right faults. A new technique digitizes the target layer. The digital layers will guide seismic attribute extraction and log propagation. After synthetic seismic, the time-domain logs from vertical and/or deviated borehole can be inserted into seismic canvas, and well markers also can be displayed along well logs.

Seismic Time-Frequency Attribute Designed to decompose post-stack seismic dataset into time-frequency panel through an implementation of Match-Pursuit Wigner-Ville Distribution (WVD). The time-frequency panel will be used to propagate logs.

Seismic Time-Velocity Attribute Designed to transform pre-stack seismic gathers into time-velocity panel through an implementation of Fast Hyperbolic Radon Transform, which has been used to de-multiple in seismic processing. The time-velocity panel will keep most of seismic gather information including amplitude versa offset (AVO). A mute window will be designed to remove multiples and other noises

Seismic Attribute Analysis The Factor Analysis will be used to classify or reduce the attributes from post and/or pre-stack seismic dataset. Usually Factor Analysis can be applied in order to explore a content area, structure a domain, map unknown concepts, classify or reduce data, transform data, define relationships, etc. R-Factor Analysis involves factoring variables, which will reduce attribute space from a larger number of variables to a smaller number of factors. After R-Factor analysis, the post-stack attributes and/or pre-stack attributes will be reduced to produce the HYBRID attributes, which is a combination of the optimum set of attributes. The HYBRID attributes have better correlation to well data.

Seismic Anisotropic Advisor (seisSuperMan) Both seismic survey and well logs are Cartesian (x, y) coordinate system. However the some reservoir properties have preferred orientations. To study the anisotropic effects, the Support Vector Regression should be performed in the flow-oriented (s,n) coordinate systems. Usually the anisotropy settings include a ratio and angle setting. The ratio is the maximum range divided by the minimum range; the angle     () is the orientation (s direction).

Seismic-Guided Log-Propagated Property (seismic geostatistical  inversion):  (seisSuperMan) Designed to propagate time-domain logs, such as porosity and water saturation, which utilizes time-frequency attributes from post-stack, time-velocity and time-radon attributes from pre-stack seismic. The state-of-the-art technology, Support Vector Regression (SVR), will be used to build the bridges between time-domain logs and seismic hybrid attributes. Anisotropic advisor will support multi-sector and multi-section analysis for logs propagation. During SVR mapping phase, the log-propagated property (named SeisLog) will be generated. The SeisLog properties, such as porosity, have the same vertical resolution as time-domain logs and the property is a better fit with borehole logs.