Currently

the most widely used pre-stack simultaneous inversion algorithm is based on the Zoeppritz equation approximation of isotropic horizontal media. This method

however

encounters challenges in medium-basic volcanic reservoirs characterized by rapid lateral lithological changes and vertical multi-period stacking. In such environments

tuffs and sedimentary tuffs exhibit similar logging and geophysical responses

complicating the distinction of lithology and physical properties of volcanic rocks through pre-stack simultaneous inversion. To address these challenges

qualitative analyses were conducted on the seismic response characteristics of volcanic reservoirs using convolutional model forward modeling. Additionally

well rock physics was employed for physical analyses of lithology and sensitive parameters pertinent to volcanic reservoirs. Comparative analyses of six binomial and trinomial pre-stack inverse algorithms

aimed at approximating the Zoeppritz equation in the specific area

were conducted through model calculations and actual data assessments. Based on these comprehensive analyses

the approximate algorithms developed by SMITH GIDLOW and FATTI were identified as preferable choices. These algorithms use the reflection coefficients of P-wave impedance

S-wave impedance

and density as inputs for pre-stack inversion. The inversion results for P-wave impedance were utilized to predict the presence of local tuff

while the inversion outcomes for density were employed to ascertain the effective reservoir physical properties. The efficacy of this approach was validated through the deployment of an evaluation well and a horizontal well

which yielded predicted compliance rates of 76.0% and 84.6%

respectively.