Enhancing temporal resolution in ground penetrating radar data through sparse blind deconvolution with the Smoothed One Over Two (SOOT) algorithm and Gabor Deconvolution (GD)

• Published in: Journal of applied geophysics Vol. 241; p. 105833
• Main Authors: Moghaddam, Sadegh, Goudarzi, Alireza, Dehghan-Manshadi, Seyed Hadi
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.10.2025
• Subjects: Compression; Gabor Deconvolution (GD); Ground Penetrating Radar (GPR); Majorization-Minimization (MM) algorithm; Smoothed-One-Over-Two (SOOT) algorithm; Sparse Blind Deconvolution (SBD); Compression; Majorization-Minimization (MM) algorithm; Ground Penetrating Radar (GPR); Smoothed-One-Over-Two (SOOT) algorithm; Sparse Blind Deconvolution (SBD); Gabor Deconvolution (GD)
• ISSN: 0926-9851
• DOI: 10.1016/j.jappgeo.2025.105833

We present an innovative approach for Sparse Blind Deconvolution (SBD) of Ground Penetrating Radar (GPR) data, employing the Smoothed One Over Two (SOOT) algorithm. This methodology achieves the restoration of underground reflection sequences without relying on prior information from the GPR pulse. The deterministic cost function of the SOOT method is formulated through a smooth approximation of the (L1/L2) norms, with the primary objective being the enhancement of resolution in particularly thin underground layers, guided by the sparsity assumption. Our investigation involves a comprehensive analysis of the Majorization-Minimization (MM) algorithm and the efficiency of the Gabor Deconvolution (GD) in generating optimal reflectivity series, serving as the core values in the SOOT algorithm and ensuring well-compressed outcomes. Results indicate that, compared to the MM algorithm, the GD method provides superior initial reflectivity sequences for the SOOT algorithm. This, in turn, enhances temporal resolution. Compared to the MM algorithm, the GD method adeptly promotes events, preserving the signal and facilitating optimal compression. The successful application of the GD method in generating initial reflectivity sequences for the SOOT algorithm leads to a notable improvement in temporal resolution. •Blind Deconvolution enhances GPR data for better results.•SOOT algorithm yields satisfactory outcomes in applications.•Study compares Majorization-Minimization and SOOT algorithms.•Gabor deconvolution excels in providing initial reflectivity.•Combination of SOOT and GD algorithms enhances GPR resolution significantly.