Mapping gas seeps with multibeam water column data in the Norwegian offshore – Challenges and way forward

Seepage of gases from natural sources of subsurface hydrocarbon accumulations escaping through the seafloor to the water column is recorded in many parts of the world’s oceans. These occurrences can be acoustically and visually observed using various sensors onboard various platforms. This is particularly common when carrying out bathymetric surveys of large areas using multibeam echosounder systems with the capability of recording the whole water column acoustic backscattering. The so-called “water-column data” that these systems produce can then be inspected for acoustic anomalies that are characteristic of gas seepages (i.e. acoustic “gas flares”), and those instances and their attributes (e.g. strength, confidence, height, etc.) can be recorded in a database. The Geological Survey of Norway has been building such a database for the Norwegian offshore since 2010. To date, this database includes over 5,000 flares of varying magnitudes and sizes, detected in an area of >140,000 km². The water-column data used for this task mainly originates from the many multibeam surveys carried out since 2005 over large areas of the Norwegian offshore for the MAREANO program, which is aimed at mapping habitats, but also from datasets acquired in associated projects and sources.

We present the results from these comprehensive surveys and discuss the various challenges faced in making such a database. Our main challenges are the very large size of the datasets and our reliance on visual interpretation, which necessitate dedicated software, high-performance processing systems, storage solutions of very large capacity and fast access, considerable interpretation time, and procedures of cross-validation between different interpreters. Another challenge is the variety of the data and its quality due to various acquisition parameters, weather conditions, and water depths, but also from the use of various systems, models, generations, and frequencies. This variety impacts the visual aspect of acoustic gas flares and thus affects the ability of the interpreters to consistently estimate flare magnitude and size. However, this variety also presents research opportunities. For example, we possess several instances of acoustic gas flares that were imaged with a range of frequencies, allowing for frequency dependence analysis. Finally, we will discuss future possibilities for interpreting water-column data in more time-efficient and interpreter-independent manners.