The middle and deep potassium salt reservoirs in the western Qaidam basin are closely related to sedimentary water changes. High- density two- dimensional seismic data, with its rich frequency information, helps in identifying sedimentary environments and provides support for the study of deep potassium- rich brine reservoirs. Continuous wavelet transform time- frequency analysis extends seismic signals from the one- dimensional time domain to the two- dimensional time- frequency domain. This allows for a can more clearlyer characterization of the cyclic structural characteristics inside within sedimentary bodies and their lithological associations in the frequency spectrum. Based on forward modeling data and actual drilling data, the corresponding seismic time- frequency signature of sedimentary bodies were concluded under the sequence stratigraphy framework(low- stand system tract (LST), transgressive system tract (TST), high- stand system tract (HST) and regressive system tract (RST)). The time- frequency energy of LST is mainly concentrated at low frequencies, while high- frequencies for HST . The time- frequency energy of TSTdecreases over time and shifts toward a higher frequency and conversely, for RST we identified 35 Hz as a key frequency threshold for dividing favorable sedimentary cycles, marking the transition between the end of RST, and the early stage of LST or HST. Selecting an appropriate frequency range for seismic data reconstruction can enhance the reflection characteristics of low- amplitude sandstone. Combining this with full- frequency band data facilitates the rapid identification of favorable sedimentary cycles, providing a valuable basis for exploring gravel- type deep brine potassium deposits.