| Abstract Submission No. | ABS-4.1-0070 |
| Title of Abstract | Coupling AUV-Based Geophysical Mapping and Image Classification to Characterize Cold Seeps and Deep-Sea Ecosystems in the Krishna¿Godavari Basin |
| Authors | Lasya Priya*, Sriram G, Komal Rani, Dewangan P, Ramesh S, Ramesh N.R |
| Organisation | CSIR-NIO |
| Address | Nio Panjim, Goa, India Pincode: 403004 E-mail: challalasyapriya2002@gmail.com |
| Country | India |
| Presentation | Poster |
| Abstract | Autonomous Underwater Vehicles (AUVs) enhance deep-sea exploration by enabling high-resolution mapping of cold seeps, gas hydrates, and vents. In December 2023, NIOT, CSIR¿NIO, and NCPOR jointly surveyed the Krishna¿Godavari Basin, following CSIR¿NIO¿s 2018 discovery of an active cold seep. The AUV, equipped with multibeam bathymetry, water column imaging, synthetic aperture sonar, sub-bottom profiler, and HD cameras, operated at 30 m and 5 m altitudes, achieving ~1 m bathymetric and ~25 cm photographic resolution. Over a 500 × 180 m area, it mapped 100+ pockmarks and ~303 gas flares from WCI data, revealing seabed morphology and fluid escape features. These align with active seepage zones linked to carbonate deposits, shallow gas hydrates, and widespread chemosynthetic communities. The survey revealed extensive faunal distributions in areas with strong gas venting seen in MB and WCI data. To better understand these communities, pixel-based image classification was done using ilastik, a toolkit usually used for microscopic images but adapted here for deep-sea photographs. This pixel-level segmentation enabled systematic classification of seafloor habitats and biota, leading to the identification of nearly ten distinct features. These include Bathymodiolus sp., goose barnacles (Neolepas) attached to Bathymodiolus sp., bacterial mats, sea spiders, dead Bathymodiolus specimens, Calyptogena sp. clams, tube worms, faunal burrows, Siboglinidae tubes, and backscatter gas bubbles. The ability to identify both live and relict chemosynthetic communities highlights the potential of combining AUV imaging with advanced computational classification for ecological assessments in deep-sea environments. This study shows the effectiveness of AUV-based surveys in detailed characterization of cold seep ecosystems. By integrating geophysical data, water column measurements, and high-resolution imagery with advanced image processing, this study establishes a framework for seep studies. The results offer essential baseline data for environmental monitoring, resource assessment, and understanding fluid dynamics, biogeochemical cycles, and benthic ecosystems in deep-sea settings. |
| Keywords | AUV, Gas Hydrate, Cold Seep, Chemosynthetic Communities, Image Classification |