UBC Okanagan researcher Saeed Mohammadiun’s work involves reviewing marine oil spill management including how computational techniques based on real-time data can be applied to an effective oil spill response.
The eyes of the world turned to a California beach recently as an underwater pipeline ruptured, leaking more than 550,000 litres of oil into the waters off the coast.
A large oil spill is one of the worst nightmares for environmentalists and local people living in an affected area, says UBC Okanagan researcher Saeed Mohammadiun. His work at UBC Okanagan’s Life Cycle Management Lab involves reviewing marine oil spill management. The School of Engineering doctoral student examined a decade of past management responses to oil spills. His most recent research looks at computational techniques based on real-time data and how they can be applied to an effective oil spill response.
What are some of the key takeaways from your latest paper when considering the recent marine oil spill in California?
Oil spills are tragic events and may have catastrophic environmental, human health, and socio-economic consequences; in California, local authorities are calling it an “environmental catastrophe.”
Even small-scale incidents may concern stakeholders, such as the recent incident in Bella Bella, BC, which has greatly affected the socio-economic status of the local indigenous community. The spilled oil may cause drastic consequences for local economies such as tourism and fisheries.
One essential step to minimize consequences of oil spills is to enhance the response preparedness by employing advanced scientific techniques such as remote sensing and data mining. Intelligent computational techniques can be used along with advanced detection and monitoring methods, such as remote sensing, to facilitate an effective and timely oil spill response. We have critically reviewed recent articles in this field and suggested a holistic framework for effective management of oil spills.
How does marine oil spill management (MOSM) minimize the impacts of oil spills?
MOSM can broadly cover multiple components such as oil spill detection and monitoring, risk assessment, response method selection and process optimization, and waste management.
An effective MOSM, based on appropriate computational-technique-based tools, can consider both proactive and reactive practices for oil spill prevention and also mitigate adverse impacts of an oil spill if it happens.
The three major benefits of an effective MOSM are the ability to detect oil spills in a timely manner, conduct the most appropriate oil spill response and reduce amounts of oily waste generated from response operations. An effective MOSM ultimately leads to minimizing the impacts of spilled oil.
Describe the role of robust computation techniques based on real-time data in reducing oil spill impacts
One important component of MOSM is the response operation, which aims to contain and clean up the spilled oil and contaminated area before any significant adverse environmental and economic impacts can occur. Response to an oil spill is a time-sensitive and complex task that is significantly reliant on available data, resources and decisions made.
Another significant challenge is the fact that collected oily wastes can be up to 10 times more than the original oil spill volume. Oily waste management is usually the bottleneck of all clean-up operations due to limitations in local resources including storage infrastructure and transportation facilities.
Inappropriate management of generated oily waste may exacerbate the situation because of secondary contamination. The application of intelligent computational techniques is essential to help stakeholders make timely and suitable decisions based on limited data. A real-time computational-based tool should be able to systematically find the most appropriate management practices by taking a considerable number of influential factors into account.
You reviewed similar research for the past 10 years. What reoccurring themes did you discover?
The risk of oil spills has increased by growing marine oil exploration and transportation activities. This trend can also be seen in arctic and sub-arctic waters under the effect of climate change, such as the Canadian Arctic, which has seen a significant increase in ship traffic and oil exploration activities in recent years. It is estimated that about two million tons of oil enter the marine environment every year around the world.
Application of artificial intelligence and soft computing techniques in oil spill management has been a new trend in this research area due to better data availability and advancement in computational techniques. Better meteorological and oceanic data, better satellite image access, advances in oil spill trajectory simulators as well as progress in artificial intelligence and soft computing techniques have collectively increased the number of studies in this field. There is still a dire need for better real-world oil spill data to enhance the performance of intelligent decision-making tools.
What role can artificial intelligence and machine learning play in minimizing the impacts of oil spills or improve oil spill response?
Both can play an important role concerning various components of oil spill management—from timely detection in remote areas to optimal response selection and waste management. An appropriate oil spill management strategy should be determined based on a significant number of factors, such as the dynamic characteristics of spilled oil and environmental conditions. Remote sensing methods can be integrated with artificial intelligence techniques, analyzing features of images, to accurately detect and monitor oil spills in offshore and onshore regions.
After the detection, it is usually necessary to respond to an oil spill immediately. To this end, artificial intelligence and machine learning techniques can considerably facilitate the decision-making process by analyzing previous oil spill data and subtle patterns that are often hidden in the data.
These intelligent tools facilitate the instant selection of the most suitable response method and its operational process.
In other words, machine learning can use previous oil spills to help manage future incidents. Also, the volume of generated oily waste can be minimized by the application of intelligent computational techniques, because the optimal oil spill response also takes waste management into the consideration.
Mohammadiun’s research was published recently in the Journal of Hazardous Materials. The paper is part of the Multi-Partner Research Initiative (MPRI), supported by Fisheries and Oceans Canada. The MPRI aims to help the oil spill response industry and regulators to enhance response preparedness in Canadian waters.
