- Research Article
- 10.1088/1755-1315/1590/1/012007
Reservoir Characterization Using Integration of and Model-Based Seismic Inversion and Neural Network Method, “M” Field, East Java Basin
- Feb 01, 2026
- IOP Conference Series: Earth and Environmental Science
- Haiqa Sabrina Ramadhani + 2 more +2
Abstract Seismic inversion is a key technique in reservoir characterization, enabling the transformation of seismic reflection data into quantitative rock property models such as acoustic impedance. This process is essential for enhancing gas field productivity by identifying reservoir features and estimating critical rock parameters. However, conventional inversion methods often struggle to accurately predict complex subsurface properties like porosity and density due to limitations in seismic resolution and the nonlinear nature of rock physics. To overcome these challenges, this study introduces an integrated methodology that combines acoustic impedance-based seismic inversion with Neural Network-driven multi-attribute analysis to evaluate a gas reservoir in the M Field, East Java Basin. Utilizing high-resolution pre-stack seismic data, the approach estimates the spatial distribution of porosity, density, and velocity. Additionally, the Neural Network analysis is employed to assess a shallower, less-explored interval, characterized by acoustic impedance values between 2948.46 and 3208.68 (g/cc*m/s), porosity levels of 18-20%, sandstone-like density ranging from 1.80 to 1.83 (g/cm 3 ), and Vp/Vs ratios from 1.83 to 1.77. This integrated workflow enhances reservoir prediction accuracy and deepens the understanding of gas-bearing zones in the study area.
Read more