AVO
    fluid inversion is a program which combines AVO theory with statistics, discriminates
    the AVO response between actual seismic data and log data, and predict
    quantificationally the probability of hydrocarbon in reservoirs. To obtain the
    quantificational porosity prediction of low-porosity and low-permeability
    reservoirs at target area, the authors present an improvement of AVO fluid
    inversion to make full use of its advantages in statistical analysis with
    probability theory such as Bayes’ theorem and Cauchy distribution, to
    distinguish anomaly information and extract effective information about
    reservoir porosity. The results in X69 area show that 79% of the prediction fit
    to the calculated values of logging data, and the distribution of effective
    reservoirs in high productivity area is accurately delineated.