Nonlinear inversion of electrical resistivity sounding data for multi-layered 1-D earth model using global particle...

Area of Science:

• Geophysics
• Computational Science
• Earth Science

Background:

• Geophysical data interpretation often involves solving complex nonlinear optimization problems in inversion.
• Traditional analytical methods like least-square inversion face limitations such as slow convergence and dimensionality issues.
• Heuristic-based swarm intelligence algorithms, particularly Particle Swarm Optimization (PSO), present a promising alternative for tackling these challenges.

Purpose of the Study:

• To evaluate the effectiveness of Global Particle Swarm Optimization (GPSO) for inverting geoelectrical resistivity data.
• To compare the performance of GPSO inversion against the conventional least-square inversion method.
• To assess the accuracy and convergence properties of GPSO for constructing 1-D multi-layered earth models from Vertical Electrical Sounding (VES) data.

Main Methods:

• Development and application of a Particle Swarm Optimization (PSO) algorithm for geoelectrical data inversion.
• Inversion of field Vertical Electrical Sounding (VES) data to determine a multi-layered 1-D earth model.
• Comparative analysis of GPSO inversion results with those obtained from the least-square inversion method using Winresist 1.0 software.

Main Results:

• GPSO achieved satisfactory inversion results with fewer than 200 particles and convergence in under 100 iterations.
• GPSO demonstrated a significantly lower misfit error ($$) compared to least-square inversion (4.0).
• The GPSO inversion model, incorporating parameter limits, provided a better fit to the true earth model, yielding more accurate inverted models than least-square methods.

Conclusions:

• Global Particle Swarm Optimization (GPSO) is an effective technique for inverting geoelectrical resistivity data, offering improved accuracy and faster convergence over least-square methods.
• While PSO inversion exhibits slower execution times and requires prior knowledge of the number of layers, its ability to estimate models closer to true solutions is a significant advantage.
• The study validates GPSO as a robust tool for geophysical inversion, particularly for complex, large-scale nonlinear optimization problems in earth modeling.