Abstract：According to geological and seismic data,we design a series of models depending on differential coal thickness. These models simulate the time profiles of the sub-waves of different frequencies,and analyze the extracted properties. Then,the attributes and variation law associated with the coal thickness could be summarized. When Coal thickness is within the range of λ /4,most of the attributes are generally witnessed as the increase of a linear relation with coal thickness,while the amplitude shows the decreased linear relations; when coal thickness is between λ /4 ~ λ /2,the relationship of amplitude and coal thickness variation is not obvious when coal thickness is beyond λ /2. Finally,we choose a seismic attributes-based multivariate regression analysis to predict coal thickness in Yima,Henan province,and have received a good effect.

Keyword：coal thickness;forward modeling;seismic attribute;regression analysis;