Characterization of seafloor pockmark seepage of hydrocarbons employing fractal: A case study from the western continental margin of India

Here we apply quantitative technique to describe the seafloor seepages based on the multi-beam backscatter and bathymetric investigations to characterize the pockmark morphology. The variable seafloor backscatter strength for coarser seafloor sediments are related to the diagenesis derived from biodegraded seepages. In this regard, box counting method is used to estimate 'fractal dimension' for backscatter imagery data of 398 blocks. These blocks are further sub-grouped into six classes depending on the spread of pockmark related seepages. The study area lies 102 km west off Marmagao along the central west coast of India which contains pre-dominantly (70%) gas-charged sediments. Comparison between the estimated self-similar fractals reveals that there is approximately 97% correlation between the box (D(box)) and information (D(info)) dimensions. Box dimension derived fractal dimension values, suggest that the seepages are more along the fault trace in deeper waters, in comparison to sparsely distributed shallow water seepages. Besides, this poor seepage is confined within the smooth to moderately rough seafloor. It is established that the high backscatter strength along the upper slope of the pockmark region having higher fractal dimensions reflects multifractal behavior of seepage distribution. Entire area indicates patchy seepage patterns as supported by estimated fractal values showing intermittent fluctuations, which emphasizes non-linear behavior. Estimated self organizing criticality (SOC) parameters for six representative blocks reveal that the nature of pockmark, fault trace, sediment nature coupled with slumping of pockmark's wall, sediment movement due to bottom currents are controlling the dynamic balance in the area seepage system. Further, our study emphasizing the multifractal behavior of seepage blocks, clearly depicts the drift in the seepage pattern. (C) 2011 Elsevier Ltd. All rights reserved.