College of Geo-Exploration Science and Technology��Jilin University��Changchun 130026��China

Making a comprasion of the analog signal contrast of Wigner--Ville distribution with those of smooth Wigner--Ville distribution��it is shown that the smooth Wigner--Ville distribution can adjust the window length�� which beacomes an effective method to eliminate the cross terms�� For reservoir prediction based on the effective extraction with the low-frequency information��we use the smooth Wigner--Ville spectral decomposition technique for prediction of oil and gas reservior��and by making seismic waves of high energy attenuation and low-frequency energy enhancement features�� Then we apply the smooth Wigner--Ville distribution in the working area for post-stake data processing��through the different frequency slice response of spectrum decomposition��and get the abnormal phenomenon of low frequency in target layer�� The results show that based on high resolution smooth Wigner--Ville distribution�� can point out thin layer better��