To optimize operations at the train derailment site
ENVIRONMENTAL MONITORING
Monitoring rehabilitation work – Lac-Mégantic
Project Start:
July 2013
Length:
4 months
The train accident occurred on July 6, 2013, at 1:14 p.m. EST in Lac-Mégantic, a municipality in the Eastern Townships, Quebec, Canada. The derailment of a convoy of 72 runaway tank cars containing light crude oil caused explosions and a fire that destroyed about 40 buildings in a 2 km2 area in the downtown, at the same time killing 47 people. The fire lasted four days, consuming nearly six million litres of oil.
Context
Since the site was declared a crime scene, the Sûreté du Québec monitored it 24/7 for several days. It was only five days after the tragedy that rehabilitation companies were able to access the site, and a few days later, started moving the cars and pumping the oil that had spilled into the water and removing the contaminated water. The off-gassing, heat and humidity made the job difficult for the rehabilitation team.
Challenges
We had to provide a series of clear and very high resolution satellite images, in spite of changing atmospheric conditions to monitor the progress of rehabilitation operations on the site over the four months following the train accident.
Although all the images were acquired using sensors (on the Pléiades 1A and 1B satellites) with identical characteristics, weather conditions, which varied greatly from day to day, resulted in the same target on the ground appearing a slightly different colour. Therefore, images had to be corrected to reduce or eliminate the atmospheric effects, and we had to not only assure an aesthetically pleasant product but also eliminate interpretation errors that could occur due to the colour differences. Moreover, it was also necessary to fix geometric distortions in the images so that they could be perfectly superimposed.
Although all the images were acquired using sensors (on the Pléiades 1A and 1B satellites) with identical characteristics, weather conditions, which varied greatly from day to day, resulted in the same target on the ground appearing a slightly different colour. Therefore, images had to be corrected to reduce or eliminate the atmospheric effects, and we had to not only assure an aesthetically pleasant product but also eliminate interpretation errors that could occur due to the colour differences. Moreover, it was also necessary to fix geometric distortions in the images so that they could be perfectly superimposed.
Solutions
The constellation of the high-resolution Pléiades satellites was programmed to acquire new images between the time of the explosion on July 6, 2013, and November 4, 2013. The images were processed using image analysis software to make them perfectly superimposable and balanced in terms of colours.
Results
As the images were obtained and corrected, they were able to be used by organizations heading up the coordination of the on-site work. They thus had access to an overview of the affected site, and, in addition to directing and optimizing the rehabilitation team efforts, could also document work progress on the Lac-Mégantic site over the four months following the accident.
Key takeaways
- Ability to acquire very high-resolution satellite images (50 cm) at a high frequency and based on requirements to monitor developments in this large-scale job
- Quick processing of images using analysis software to facilitate the extraction of useful information.
- Supporting teams in the field and decision makers in order to have an overview of the damage caused in the areas affected and monitor developments in rehabilitation activities.
