Document Type : Original Research Paper
Authors
1 Department of Geophysics, Ha.C., Islamic Azad University, Hamedan, Iran.
2 School of Mining, Petroleum and Geophysics, Shahrood University of Technology, Shahrood, Iran
Abstract
Accurate modeling of gravity anomalies caused by geological structures remains a central challenge in exploration geophysics. In mineral exploration, geophysical targets can be approximated using simple geometric shapes, and inversion techniques are employed to estimate the parameters of buried anomalous bodies. In this study, gravity data from the Cham Gilan bitumen deposit were interpreted using Improved Particle Swarm Optimization (IGPSO) and Levenberg–Marquardt (LM) algorithms. According to borehole information, the mineralized zone was represented by a dipping slab. To model gravity anomalies with these approaches, a two dimensional inclined slab of finite extent was first considered as a synthetic model. Using the metaheuristic IGPSO and deterministic LM algorithms, five parameters of the 2D slab (dip angle, upper depth, lower depth, thickness, and density contrast) were estimated. Synthetic gravity data were generated under two conditions: noiseless and with 5% Gaussian noise. Each inversion algorithm was executed 30 times (Monte Carlo runs) to evaluate accuracy, stability, and computational efficiency. Dispersion analysis indicated that LM, due to its deterministic nature, produced near zero standard deviation for noiseless data, while IGPSO exhibited notable stability despite its stochastic characteristics. The average computation time for LM and IGPSO was about 0.04 s and 1.44 s, respectively. Both methods were further employed to interpret the gravity data of the Cham Gilan bitumen zone, and the inversion results showed strong agreement with geological observations and core drilling evidence.
Keywords
Main Subjects
)