Well log normalization can effectively reduce systematic errors during logging data interpretation

therefore has been used to improve the accuracy of formation parameter calculation and reservoir prediction.However

lack of a stable marker bed makes the marker-based normalization method too rigid to be adapted to continental sedimentary sequences with rapid variation of lithology.This paper analyzes the geological meaning of spectral decomposition

determines the contributions of different cycles to sedimentary formations

and then regards the superimposed formation of low-frequency cycles which represents the background values of formation as a macro marker bed with stable lateral distribution.The extreme value of low-frequency component was selected as characteristic value and the marker bed values were determined through trend surface analysis.Then translation correction and scale correction were carried out on the low-frequency and high-frequency components of the curves.Finally

log normalization can be achieved.This method had been applied to the normalization of logs of the Putaohua Formation in Daqing central uplift in Songliao Basin

and the correlation coefficient of core porosity and normalized curves was significantly higher than that of core porosity and original curves.This demonstrates that the log normalization method based on spectrum decomposition can be effectively applied to the continental sequences without stable marker beds.