Abstract: Aiming at the difficulties that some high-porosity water-bearing sandstones and oil-bearing sandstones in deep-water clastic reservoirs in West Africa have similar amplitude bright spots and show Class III AVO anomalies with remote-trace amplitude enhancement,this paper proposes to use flat spots to detect the hydrocarbons in the reservoirs.As revealed by the data on the geometric shapes,amplitude,and phase of seismic reflection,four types of flat spots are have developed in the study area,namely individual flat spots,short-axis double flat spots,short-axis composite individual flat spots,and long-axis composite individual flat spots.The target layer W is dominated by short-axis double flat spots.Based on the forward modeling with reflection coefficients and stratigraphic dip as core parameters,a quantitative discrimination template of seismic reflection of flat spots was established in this study.The flat spot distribution was determined using the flat spot strengthening technology and the common isoline trace gathering stacking technology.It was inferred from the discrimination criteria of hydrocarbon flat spots that the upper flat spot strips represent gas-oil interfaces and the lower flat spot strips represent oil-water interfaces.Based on this and the analysis of sensitive attributes in hydrocarbon detection,the area of oil- and gas-bearing zones was effectively predicted.Therefore,the application of flat spots can reduce the multiplicity of solution in the hydrocarbon detection and improve the prediction accuracy of reservoirs.