Overview

Strong metocean conditions, such as extreme ocean currents and natural shifts in the ocean seabed, can have a significant impact on the integrity and longevity of offshore infrastructure. Changes in the ocean environment, such as submarine landslides, can be triggered by strong weather events (e.g., hurricanes) as well as the natural fluctuations in the system over time. These changes can impact safe operations and use of offshore infrastructure, and metocean events to date have been linked to billions of dollars in damage and environmental and community impacts. As a result, the ever-shifting offshore environment introduces a currently unquantified level of risk and vulnerabilities to a range of offshore energy activities.

This project will leverage historic metocean and bathymetric conditions to develop a new smart technology that will help identify current hazardous metocean and bathymetric conditions, as well as forecast changes and potential vulnerabilities that may affect existing or future offshore infrastructure and activities. The result will be an artificial intelligence/machine learning (AI/ML) and probabilistic smart tool that advances the current state of knowledge of ocean and seabed conditions and offshore hazards. This smart tool will offer insights to improve infrastructure longevity and support the identification of seabed and shallow subsurface hazards, thus mitigating environmental impacts, reducing costs, and improving safety during engineered-natural system energy activities.

Approach

To successfully develop, train, test, and validate a smart tool requires a significant volume, and often velocity, of data to support the analytics. However, the multi-variate, multi-dimensional, and often restricted nature of metocean and bathymetric data has limited the ability of researchers to support these robust analytics until more recently due to the development and broad application of new sensors, sonars, and big data processing techniques.

This project will facilitate the development, training, testing, and validation of a smart technology that uses historic observations of metocean and bathymetric data to identify current hazardous metocean and bathymetric conditions, as well as forecast changes and potential vulnerabilities that may impact existing or future offshore infrastructure. The smart technology will integrate AI/ML, spatio-statistical, and probabilistic analyses, where appropriate, to accelerate analyses and provide insight into potentially hazardous conditions.

Year 1 research focused on proving the concept — that sufficient, detailed historic data are available to support the development of an analytical framework that uses advanced analytics (such as machine learning, artificial intelligence, and big data capabilities) to forecast and predict metocean and bathymetric hazards and vulnerabilities. In Year 2, the smart tool framework, known as the Ocean & Geohazard Analysis (OGA) tool, was developed, and multiple AI/ML and/or probabilistic analytical models for corresponding metocean and seabed hazard conditions were developed to predict vulnerabilities. Year 3 will focus on validating and refining the analytical logic and interactive, online tool based off input from internal and external stakeholders. Moving into Year 4, we will continue to optimize the OGA tool and publicly release an integrated metocean and seabed hazard database.

Expected Outcome

This project will result in a smart, interactive, online tool to predict potential metocean and bathymetric hazards and vulnerabilities in relation to improving infrastructure longevity, shallow hazard reduction,

and offshore energy activities. The tool will operate as a plug-in application within EDX®’s Geocube, enabling the ability to perform near-real time assessments of current hazards and the support of forecasting areas with higher vulnerabilities. In addition to the tool, key analytical results will also be made available as interactive map layers. Together, this tool and information will aid in the reduction of potential risks to daily offshore activities including carbon storage and offshore wind/wave energy in the future.

Research Products

Research products are being developed and will be posted here as updates when publicly available.

Explore research products that are related to this project.

Figure 1. CIIAM outputs for ocean current trajectory showing how tracers representing hazmat material would deform under the influence of ocean currents in the Gulf of Mexico.

Figure 2. Predicted seafloor landslide potential in the northern Gulf of Mexico using spatial risk-based analysis.

Figure 3. Predicted seafloor landslide potential in the northern Gulf of Mexico using AI/ML-informed analysis.

Figure 4. Automated landslide detection results from a data-driven neural network analysis of northern Gulf of Mexico bathymetry. Recent results (left) show the two output layers of the network (bottom C, D) which are combined to create the prediction mask (A) with the ground truth (B) for comparison.

Figure 5. Temporal variability of sea-surface velocity obtained objectively through self-organizing maps, from the western equatorial Atlantic to the Gulf of Mexico, showing a seasonal cycle in the region where Loop Current eddy separation initiates.

Figure 6. Significant wave height for the 100-years return period obtained from physics-based modeling of the GCM derived events ensemble for the (a) present and (b) future wave climates. Blank areas denote regions where less than 4 models show the same trend.

Contacts

Jennifer Bauer
Co-Principal Investigator

Rodrigo Duran
Research Applied Mathematician/Oceanographer
Co-Principal Investigator

MacKenzie Mark-Moser
Research Geologist
Co-Principal Investigator

Kelly Rose
Offshore Portfolio Lead

Roy Long
Offshore Portfolio Technical Manager
Effective Resource Development

Grant Bromhal
Senior Fellow
Geological & Environmental Systems

Philip Reppert
Associate Director
Geological & Environmental Systems