Abstract: North Ustyurt Basin is a continent polycyclic-cratonic basin, its Jurassic strata developed several good sets of source rocks. Because of low exploration activity and unclear understanding of the lithofacies palaeogeography of the study area, the evaluation for source rocks is facing great difficulties. In order to solve this problem, a method using seismic attributes technique to predict mudstone content is presented, based on the analysis of regional geological background of the study area. 11 kinds of seismic attributes are extracted by Landmark according to the seismic data in exploration area. Through analysis of the correlation and sensitivity of the seismic attributes, three attributes such as RMS Amplitude, Arc Length and Maximum Trough Amplitude which are most sensitive to mudstone content are selected as the analysis parameters of prediction models of mudstone content. Accordingly, by the theory of multiple linear regression, we put forward a model predicting the content and distribution of the mudstone of Jurassic in the study area. After verification, the prediction from the model matches well with the drilling data. Finally, the favorable mudstone zones in the exploration area is indicated, there is no doubt that this study will provide reliable foundation for further exploration in the study area.

Key words: seismic attributes, multiple linear regression, mudstone prediction, Jurassic, North Ustyurt Basin