The earth absorption effect is an important factor resulting in the energy attenuation of seismic waves and the degradation of resolution. The inverse Q filtering method helps to restore attenuated energy and improve data resolution. The conventional inverse Q filtering method employs a gain function, which can adaptively reduce the compensation coefficient of high-frequency energy but cannot adapt to the signal to noise ratio (SNR)of data very well. Therefore, the gain parameter related to the local data SNR has been specially designed and proposed based on the gain function. A gain parameter with attributes of spatial and temporal variability can better adapt to the features of data. In high SNR regions, if the chosen value of the gain parameter is high, it can adequately compensate for the high frequency components in a effective manner. In contrast, in low SNR regions, if the chosen value of the gain parameter is smaller, it can only compensate for the high frequency components in a relatively limited manner. Model testing and data processing demonstrates that the inverse Q filtering method can be well adapted to data and achieve a good compensation effect.