The delta front is always one of the focuses for oil and gas exploration.However

reservoirs in delta front commonly occur as thin sandstone layers

which cannot be recognized accurately for the low resolution of seismic inversion

hence the distribution pattern of sandbodies could not be accurately predicted

leading to the difficulties for exploration and production.Taking advantage of various reservoir modeling techniques

this paper puts forward a new methodology of joint deterministic-stochastic modeling for lithology prediction of thin reservoirs in delta front areas.It first obtains acoustic impedance through deterministic modeling

and then predicts distribution of thin reservoirs through stochastic modeling.The specific methods used are joint seismic-log inversion and Markov chain modeling.Case study in the Y block of the northern Songliao Basin shows that more accurate prediction of thin sand layers could be achieved.