Seismic refraction data inversion via jellyfish search algorithm for bedrock characterization in dam sites

Abstract

Seismic refraction method is an efficient tool for the investigation of dam construction sites. Velocity inversion has an essential role in an accurate seismic refraction data interpretation. This study aims to develop a new inversion algorithm to estimate P-wave velocity (Vp) structure from seismic refraction travel times. The introduced inversion algorithm is based on a recently developed nature-inspired algorithm, i.e., jellyfish search (JS) optimizer. First, the JS-based inversion algorithm was tested by several synthetic models in the presence of noise and without noise. Then, the performance of the applied inversion algorithm was evaluated by the seismic refraction travel times at a realistic dam construction site. The main objective of the actual data set analysis is the determination of Vp structure to find overburden thickness. The JS-based inversion algorithm in both synthetic models and actual data set shows acceptable performance. Results show three distinct seismic layers at the dam site. The velocities of the first, second and third layers, respectively, were estimated 400 m/s, 600 m/s and 1400 m/s. Also, the overburden thickness was estimated about 23 m, which was consistent with borehole data. The performance of the applied algorithm in the analyzing of actual data set was compared with the tomography interpretation method that the results revealed the efficiency of the JS-based inversion method.