Bayesian inversion method for lithologic prediction
Lithologic velocity is one of the most important reservoir parameters in hydrocarbon exploration and development.A technique for lithologic velocity prediction using the 2D seismic data,borehole,and other prior information is developed in the Bayesian framework.Within the Bayesian framework,lithologic velocity parameters are considered as random variables and are estimated simultaneously by conditioning to available data which are transformed into the form of probability.So more information,including diversified types and precision data,are utilized in the inverse process,the more certain are the result of the inference.In fact,the inverse process is on sampling from the unknown joint probability density function of the model to be inversed.Then various statistics,such as mean,variance and probability density function,can be computed through these samples.A numerical test,whose geophysical model based on Ordovician reef community in Tarim Basin,and a field case are presented to test the algorithm.Inversion results show that Bayesian inversion method we designed is reliable,and also provides information about uncertainty in the prediction.
Wentao Wang Peimin Zhu
Institute of Geophysics & Geomatics,China University of Geosciences,Wuhan 430074,China
国际会议
The 3rd International Conference on Environmental and Engineering Geophysics(第三届环境与工程地球物理国际会议)
武汉
英文
151-154
2008-06-15（万方平台首次上网日期，不代表论文的发表时间）