Seismic data calibration in terms of reservoir properties with a multivariate gaussian segmentation technique

@inproceedings{joseph:hal-04026658,
 abstract = {This paper presents a statistical calibration technique to derive geological properties from seismic traces at the reservoir level. The calibration method is based on a gaussian segmentation of the calibration points in the multivariate space generated by the geological properties and seismic attributes measured at the wells and their neighboring traces. The multivariate probability density function is viewed as a finite mixture of gaussian components. This approximation of the probability density function can be used to compute the conditional distribution of the geological properties given the measurement of a vector S of seismic attributes. Various parameters characterizing the conditional distribution can be derived to quantify the geological prediction and its uncertainty at the trace under consideration. The proposed calibration technique also allows to account for non linear relationship between the geological and the seismic attributes. The method is illustrated on the geological calibration of a 2D seismic data set.},
 address = {Washignton DC, United States},
 author = {Joseph, C. and Fournier, Fr{\'e}d{\'e}rique and Royer, Jean-Jacques},
 booktitle = {{SEG Technical Program Expanded Abstracts 1993}},
 doi = {10.1190/1.1822462},
 hal_id = {hal-04026658},
 hal_version = {v1},
 number = {12},
 pages = {285-288},
 publisher = {{Society of Exploration Geophysicists}},
 title = {{Seismic data calibration in terms of reservoir properties with a multivariate gaussian segmentation technique}},
 url = {https://hal.univ-lorraine.fr/hal-04026658},
 year = {1993}
}