Abstract:Common methods for Q-value calculation are rise-time method, spectral ratio method, spectrum modeling method, wavelet modeling method and so forth. Unfortunately, these methods are sensitive to noises and time-consuming. Hence we derive the expression of fitting function and propose a new Q-value inversion method using nonlinear parameter statistical modeling of seismic records. The statistical characteristic can be used to suppress random noises successfully and to avoid the influence of source wavelets. This new method is time saving and favours calculation of accurate interval Q-value. It is favourable to direct discovery of lithologic hydrocarbon pool.