Abstract: The fitting depth and calculated interval velocity will be inaccurate when the time-depth fitting of vertical seismic profile (VSP) data is applied to deep strata. In this study, polynomials and power functions were used to fit the time-depth relation based on VSP data from 41 drilling holes in the East China Sea Shelf Basin. The applicability of the fitting in deep strata was analyzed by analyzing the seismic velocity volume data and seismic profiles. The three-dimensional seismic velocity volume from migration has a maximum depth of 8 s in two-way travel time (TWT). The average fitting depth from the quadratic polynomial is 9.2% higher and the average interval velocity is 36.2% higher, respectively, whereas the power fitting depth and interval velocity are 28.9% lower and 35.6% lower, respectively in the deep strata (TWT=8 s), where the seismic velocity volume data shows that the average depth and interval velocity at 41 drilling holes are 18 140 m and 6 208 m/s, respectively. To solve this problem, a segmented fitting model was proposed, in which the interval velocity increase turning point of the VSP data was firstly identified, then the quadratic polynomial function was used to fit the VSP data above the turning point and the power function was used to fit the VSP data beneath the turning point. At TWT=8 s, the errors of the average fitting depth and interval velocity are reduced to 3.3% and 4.7%, respectively. The segmented fitting model shows that the average fitting depth and interval velocity are 27 516 m and 7 334 m/s, respectively, and these values basically are consistent with the observed Moho depth(TWT=11 s) in the seismic section of the study area. The segmented fitting method can significantly improve the accuracy of time-depth conversion of the strata, especially at large depths.

Key words: vertical seismic profiling, seismic velocity, time-depth conversion, fitting equation, deep strata, East China Sea Shelf Basin